2.1 AIRWAY OBSTRUCTION

The most common serious respiratory complications after UPPP are airway obstruction and oxygen desaturation. Complete and untreated airway obstruction may lead to respiratory arrest, and reintubation or emergency tracheo-tomy could become necessary and lifesaving.

Although airway obstruction may be considered a common postoperative complication after UPPP, the actual incidence of such an event is quite low. Additionally, there has been a change in the number of serious respiratory complications reported over time, with more recent studies reporting lower rates. For example, in 1989 Esclamado et al.¹⁰conducted a retrospective chart review of 135 patients surgically treated for OSA and found the incidence of respiratory complications to be 10.3%. Half of these complications were airway obstructions that occurred at extubation, and half were due to failed intubation. It is important in a discussion of postoperative management to note that the onset of all respiratory events occurred in the early phase. That is, all respiratory complications occurred during or immediately after intubation. A review in 1994 by Haavisto and Suonpaa¹¹ of 101 patients undergoing UPPP for OSA showed an incidence of airway complications of 11%. Again all cases occurred in the early phase.

In a 1998 review of 109 patients undergoing UPPP for OSA, Terris et al. reported a 5.5% incidence or early respiratory complications.⁶ This rate included one case of airway compromise, and the patient required naloxone, oxygen and suctioning of his airway. This single episode of airway obstruction occurred only 15 minutes after the patient was transferred out of the operating room.

An additional 1998 study by Mickelson and Hakim of 347 patients undergoing UPPP for OSA reported only five patients who had experienced airway-related complications, or 1.4% of all patients.⁵ Three of the five suffered shortness of breath either at extubation or immediately after in the recovery room. Two of the five patients experienced complications in the surgical ward. These included one case of pulmonary edema and one case of shortness of breath of unclear origin. No actual cases of airway obstructions were reported.

Spiegel and Raval reviewed 117 patients in 2005 for complications associated with UPPP.⁸ Ten patients were discharged the same day as the surgery. The remainder of the patients experienced a 4.3% incidence rate of respiratory complications. While most cases were due to oxygen desaturation, two patients developed airway obstruction from presumed laryngospasm at the time of extubation. Both of these patients developed postobstructive pulmonary edema within minutes of the obstruction.

Overall, the rate of airway obstruction appears to be in the range of 1.4–11%. Most importantly, obstruction events appear to occur immediately after surgery in the early phase of recovery, so it is not likely that an extended period of observation would be a useful way to prevent this type of complication.

2.2 OXYGEN DESATURATION

Oxygen desaturation is also often cited as a common complication in OSA patients undergoing UPPP. A value of less than 85–93% is typically considered significant desaturation. In the past, some surgeons have advocated postoperative ICU monitoring for fear of serious desaturation events. However, such monitoring may not be reasonable since most desaturations occur in the early phase of recovery. Additionally, more mild nocturnal desaturations may continue for weeks, since UPPP does not correct OSA immediately.¹²

The fact that most occurrences of desaturations happen relatively early in the recovery process further supports same-day patient discharge for the majority of patients. For example, in a 2006 study by Hathaway and Johnson of 110 patients undergoing UPPP, desaturations in three patients all occurred in the recovery room.⁷ In the review by Spiegel and Raval, three desaturations below 90% were reported.⁸ The first desaturation occurred in the recovery room, the second on the first postoperative night and the third on the first postoperative day. The third patient likely desaturated on the first postoperative day because he was given narcotic patient-controlled analgesia (morphine) for pain, and narcotic analgesia tends to impede respiratory function.

Advocates of extended postoperative stays after UPPP also claim the need for post-surgical oxygen supplementation on the floor as an important argument for hospitalization. However, there is some evidence that there may not be a direct relationship between oxygen supplementation and lower desaturation rates. For example, Riley et al.¹³ found no significant difference in the postoperative low oxyhemoglobin saturation (LSAT) between patients receiving continuous positive airway pressure (CPAP) and the non-CPAP group.

A patient’s preoperative Respiratory Distress Index (RDI) may also be a poor predictor of whether or not they may desaturate, as was the case in the review by Hathaway and Johnson.⁷ While two of the admitted patients had RDIs over 50 (indicating severe OSA) one patient had an RDI of 22 and still desaturated. Desaturations occurred in the recovery room, and all patients were admitted into the hospital, but none required the ICU. The authors concluded that preoperative polysomnography data such as the RDI may not predict which patients will experience desaturation in the early postoperative period.

Overall, research shows that desaturation, particularly nocturnal desaturation, is a very common complication of UPPP surgery for OSA. In fact, it should be considered more of an expectation than a complication. Giving oxygen postoperatively may or may not make a difference, so admitting all UPPP patients in order to give them oxygen would not be reasonable. Also, preoperative sleep indices may not be an accurate indicator for identifying patients who may desaturate, so it would not be practical to hospitalize all patients due to fear of this complication. The most likely scenario is that if a serious desaturation occurs, it would be in the early phase of recovery, at which point the patient should be admitted to the hospital. Otherwise, minor nocturnal desaturations are to be expected since it may take a few weeks to note improvement in a patient’s OSA.

2.3 BLEEDING

Bleeding, particularly from the surgical site, is another frequently noted complication of UPPP. The onset of bleeding tends to be biphasic with cases either occurring immediately after surgery or more than 24 hours post operation.⁸

For example, a Haavisto and Suonpaa review of 101 in 1994 had an incidence rate of hemorrhage of 14% or 14 totalcases.¹¹ Out of the 14 cases, nine required surgical management such as electrocautery and ligation while five were managed with conservative treatment or suctioning. All butone of the hemorrhages that required ­electrocautery occurredwithin 5 hours of the initial surgery during the early phase of recovery. Three cases occurred much later, at 24 hours or more post procedure. One patient reportedly experienced asystole and died.

In a paper by Esclamado et al., hemorrhage that  required return to the OR occurred in 2.2% of patients immediately post operation.¹⁰ Mickelson and Hakim reported a 1.4% incidence of bleeding, or a total of five cases.⁶ This included two episodes of epistaxis, one of tracheostomy bleeding, and a third of bleeding from the patient’s tonsils. These four cases occurred in the early phase of recovery. However, one patient began to bleed from his tonsillectomy site while out on the surgical ward and required a return to the ICU.

Spiegel’s study also illustrated the biphasic nature of bleeding events and reported seven cases of hemorrhage with time of onset ranging between 5 and 14 days postoperatively.⁸ No cases of bleeding were reported during the hospital stay.

The biphasic pattern of bleeding after UPPP appears to be a common occurrence and is reported in various studies. Since most complications occur early, and those that do not occur much later than the average period of inpatient monitoring, a short observation period of 2–3 hours should be sufficient for detecting most bleeding events.

2.4 CARDIAC COMPLICATIONS

Hypertension, arrhythmia, angina and miscellaneous other cardiac complications are also observed after UPPP, but tend to be relatively rare. Significant hypertension frequently occurs intraoperatively or immediately after surgery. In the Terris et al. study, seven of 125 patients were described as having mild hypertension (systolic pressure greater than 190 mm Hg and diastolic over 100 mm Hg) and did not require intravenous anti-hypertensive drugs or ICU monitoring.⁶ Six additional patients experienced severe hypertension and required intravenous treatment, but in all of the cases hypertension manifested in the first 2 postoperative hours.

Hypertension requiring intravenous narcotics was also noted in three cases by Spiegel and Raval.⁸ All three cases occurred in the recovery area, and two of the three patients had a medical history of hypertension. Riley et al.¹³ reported the highest rate of postoperative hypertension that required intravenous medication at 70.5%. Such elevated rates may be due to the fact that OSA patients tend to experience a higher incidence of postoperative hypertension because they have a greater prevalence of hypertension prior to surgery. Comorbid factors such as elevated Body Mass Index (BMI) and age may account for some of the association, but OSA itself probably presents an independent risk factor as well.⁶ Since most cases of postoperative hypertension occur in the early phase of recovery and can be controlled with intravenous narcotics, hypertension alone should not constitute a reason for prolonged inpatient visits of ICU monitoring.

While hypertension is the most common cardiac complication after UPPP, cases of arrhythmia and angina have also been reported. In the review by Riley et al., four out of 182 patients (1.9%) had a postoperative arrhythmia that required treatment despite a negative medical history and work-up.¹³ All reported arrhythmias were new-onset atrial fibrillations. One patient had unstable angina.

Terris et al. also noted one arrhythmia, but the patient had a prior history of premature ventricular contractions (PVCs).⁶ Haavisto and Suonpaa also reported one case of arrhythmia.¹² A review of 1004 patients by Spiegel and Raval,⁸ which included the studies mentioned above, found an average rate of arrhythmia of 0.8%, and 0.2% for angina. Clearly, these complications are very rare and monitoring during the early phase of recovery should be sufficient in detecting the vast majority of such adverse cardiac events.

2.5 PULMONARY EDEMA

Postobstructive pulmonary edema (POPE) is a much feared complication for OSA patients undergoing UPPP and is often cited as a justification for admission. One theory is that POPE, also known as negative pressure pulmonary edema, is a non-cardiogenic pathologic process in which the generation of markedly negative intrathoracic pressures that are created by forced inspiration againsta closed glottis cause a transudation of fluid into the pulmonary interstitium.¹⁴ Typically, POPE has a benign and rapidly resolving clinical course, assuming it is recognized and treated in a prompt manner.

However, studies reveal that POPE can also result in significant morbidity, with mortality rates ranging from 11% to 40%, so clearly it is of concern to physicians.¹⁴ Why POPE appears in some individuals and not others is unclear. Literature has shown that POPE is usually triggered by laryngospasm during extubation. Again, this serious complication is identified in the immediate postoperative period (early) and a prolonged period of postoperative monitoring would not seem indicated for all patients. In a study by Goldenberg et al.,¹⁴ all six reviewed cases of POPE occurred within 60 minutes after the onset or relief of obstruction.

Also, it is important to note that the incidence of POPE is quite low. In a review of 1004 OSA patients undergoing UPPP, only three cases of POPE were reported.⁸ The risk of POPE can be reduced even further by using a bite lock to prevent accidental compression of the endotracheal tube. When POPE does occur, it usually manifests in the immediate part of recovery, typically at the time of extubation, so overnight monitoring would not decrease its incidence. The use of perioperative corticosteroids may help decrease airway edema,⁶ but the most important preventive measure possible is immediate recognition and treatment of the problem in the rare circumstances that it does occur.

Laryngospasm as a cause of POPE is not unique to UPPP but can occur with any surgical procedure, though it is more likely in procedures involving blood within or manipulations of the oropharynx. That removing the anatomic site of obstruction causes an immediate and uncompensated pressure change, resulting in fluid shifts, is not logical. If that were the case, certainly individuals with OSA would be at risk of developing POPE every time they wake up and their temporary obstruction is relieved.

2.6 PAIN AND DEHYDRATION

Pain, dehydration and poor oral intake may be the most reasonable justification for overnight patient admission, though they hardly warrant observation within an intensive care setting. Physicians familiar with the UPPP procedure undoubtedly are aware of how very uncomfortable patients are postoperatively. We routinely caution patients regarding the significant postoperative pain and review pain management strategies at length during preoperative counseling.

Hathaway and Johnson reported on 90 out of 110 UPPP procedures on an outpatient basis.⁷ Of the 20 patients admitted for overnight stay, eight patients were admitted due to limited oral intake and three for nausea. Five patients had a prolonged stay despite having no specific complications. The same was noted by Spiegel and Raval⁸ – several patients had a prolonged stay due to inadequate pain control and poor oral intake.

Pain control, poor oral intake and dehydration are important issues to address, though with adequate preoperative preparation and patient selection, outpatient surgery is still recommended for the majority of individuals.