Nonintubated Group Research Articles

Nonintubated Thoracoscopic Lobectomy for Lung Cancer

To evaluate the feasibility and safety of thoracoscopic lobectomy without endotracheal intubation. General anesthesia with single-lung ventilation is considered mandatory for thoracoscopic lobectomy for non-small cell lung cancer (NSCLC). Nonintubated thoracoscopic lobectomy has not been reported previously. From August 2009 through June 2010, some 30 consecutive patients with clinical stage I or II NSCLC were treated by nonintubated thoracoscopic lobectomy using epidural anesthesia, intrathoracic vagal blockade, and sedation. To evaluate the feasibility and safety of this novel technique, they were compared with a control group consisting of 30 consecutive patients with clinical stage I or II NSCLC who underwent thoracoscopic lobectomy using intubated general anesthesia from August 2008 through July 2009. Collapse of the operative lung and inhibition of coughing were satisfactory in the nonintubated patients, induced by spontaneous breathing, and vagal blockade. Three patients in the nonintubated group required conversion to intubated-single lung ventilation because of persistent hypoxemia, poor epidural anesthesia pain control, and bleeding. One patient in each group was converted to thoracotomy because of bleeding. The 2 groups had comparable anesthesia durations, surgical durations, blood loss, and numbers of dissected lymph nodes. Patients who underwent nonintubated surgery had lower rates of sore throat (6.7% vs 40.0%, P = 0.002) and earlier resumption of oral intake (mean, 4.7 hours vs 18.8 hours, P < 0.001). Patients undergoing nonintubated surgery also had a trend toward better noncomplication rates (90% vs 66.7%, P = 0.057) and shorter postoperative hospital stays (mean, 5.9 days vs 7.1 days, P = 0.078). Nonintubated thoracoscopic lobectomy is technically feasible and is as safe as lobectomy performed with intubation in highly selected patients. It can be a valid alternative of single-lung-ventilated thoracoscopic surgery in managing early-stage NSCLC.

Critical findings of severe influenza A (H1N1) pneumonia in children

Pandemic influenza A (H1N1) causes severe pneumonia in children. The mechanism of development of respiratory failure in pneumonia patients remains unknown. This report describes clinical features of childhood influenza A pneumonia. The clinical and laboratory findings of 31 H1N1 pneumonia patients hospitalized in Iwata City Hospital from 1 October 2009 to 31 January 2010 were reviewed. Intubation and mechanical ventilation were required due to respiratory failure in eight patients, who were classified as the intubation group. Other patients without mechanical ventilation were classified as the non-intubation group. Clinical features and laboratory findings were compared between the two groups. The median age was 6.3 years (range, 3-10 years). The male to female ratio was 22:9. Clinical manifestations of tachycardia, tachypnea and cyanosis were significant findings in the intubation group at admission. Lymphocytopenia was observed in both groups. Leukocytosis with neutrophilia was the risk factor for intubation. Tachycardia, tachypnea, cyanosis and leukocytosis with neutrophilia, could be useful predictors at admission to identify high-risk influenza A (H1N1) pneumonia in children.

Implementation of a guideline to improve prescription of analgesia for adult trauma patients in an Emergency Department

Summary Objectives To improve pain management of patients presenting to the emergency department with traumatic injuries through the development and implementation of a trauma pain guideline. Methods A retrospective review of 100 Emergency Department trauma patients (50 intubated and 50 non-intubated) was conducted in June 2005. Pain management guidelines were developed by the multidisciplinary group and implemented through intensive staff education in December 2005 and again in February 2006. A second review was conducted in April–June 2006. Results There was a significant increase in the number of intubated patients receiving analgesia in the post-implementation group, 32% (16/50) compared to 68% (34/50) ( p =0.0002). Use of morphine and midazolam infusions, a specific recommendation of the guideline, increased from 16% (8/50) to 36% (18/50) ( p =0.02), and the use of sedation (other than midazolam) together with analgesia increased from 16% (8/50) to 32% (18/50) ( p =0.02). In the non-intubated group the use of analgesia increased from 86% (43/50) to 100% (50/50) ( p =0.006). Median time to analgesia decreased from 38min to 14min and the use of multimodal analgesia increased from 30% (15/50) to 61% (30/50) ( p =0.002). Conclusion The implementation of trauma pain management guidelines and associated education has resulted in improved analgesic use in trauma patients in the emergency department.

Predictors of the need for rapid sequence intubation in the poisoned patient with reduced Glasgow coma score

Aim:In patients presenting to the emergency department (ED) with significant poisoning and reduced Glasgow coma score (GCS), the decision to proceed with rapid sequence intubation can be a difficult one....

Use of Scene Vital Signs Improves TRISS Predicted Survival in Intubated Trauma Patients

The Trauma Related Injury Severity Score (TRISS) has been previously validated to predict outcomes in nonintubated, nonparalyzed trauma patients. The purpose of this study was to assess the impact of scene vital signs on predicting survival in intubated trauma patients. Our Trauma Registry of the American College of Surgeons was reviewed for all trauma patients admitted between 10/01/04 and 09/30/06, arriving by aeromedical transport. TRISS was evaluated for each patient based on their (1) scene vital signs and (2) arrival vital signs. Additionally, the "TRISS-like" score was calculated for each patient. Expected mortality for each score was measured against observed mortality. Four thousand four hundred ninety-nine Trauma Registry of the American College of Surgeons patients were admitted during the study period; 695 (15%) were transported by air; 163 patients (23%) arrived intubated; 480 arrived nonintubated. Observed survival in the intubated group was 76%. Observed survival in the nonintubated group was 100%. TRISS using scene vital signs more closely predicted mortality among intubated patients than the other scoring systems (69% versus 39% using TRISS-arrival versus 80% using TRISS-like). Scene vital signs with TRISS also resulted in fewer "unexpected" outcomes (survivors and deaths). Traditionally, patients arriving at trauma centers intubated are either excluded from the trauma registry or have their physiological score "modified" to account for pharmacologically altered respiratory rate and Glasgow Coma Scale. In intubated patients, TRISS using scene vital signs more reliably predicts survival and does so with far fewer "unexpected" outcomes than with other available scoring systems.

Selective non‐intubation of a silicone tube in external dacryocystorhinostomy

To evaluate whether silicone tube insertion is always necessary in external dacryocystorhinostomy (DCR). During external DCR for primary nasolacrimal duct obstruction carried out between January 2001 and October 2004, silicone intubation was not performed selectively if the lacrimal sac was large and the nasal cavity was not severely narrowed. Of a total of 166 DCR cases in 153 patients, no silicone tube was placed in 74 eyes of 69 patients (44.6%). Anatomic patency of rhinostomy was achieved in all non-intubation group eyes. However, four (6.7%) of these 69 eyes showed persistent epiphora even with anatomic patency. Silicone tubing can be avoided in about 50% of cases of external DCR without detrimentally affecting the success rate. Cases in which intubation was avoided had a large lacrimal sac and a wide nasal cavity.

Effets de la musicothérapie en réanimation hors sédation chez des patients en cours de sevrage ventilatoire versus des patients non ventilés

Objectives Music has been found to be an effective nonpharmacologic adjunct for managing anxiety and promoting relaxation in limited trials of critically ill patients. However, its effects have not been compared in intubated patients during weaning from mechanical ventilation with non-intubated patients spontaneously breathing. Study design A cross-over randomized experimental design. Patients and Methods Thirty patients were studied (intubated group n = 15, non-intubated group n = 15). Patients were randomized to receive either 20 minutes of uninterrupted rest or then 20 minutes of music therapy or the music therapy first and then the uninterrupted rest period. Patients selected a relaxing music of their choice from a selection including different types of music. Heart rate (HR), systolic blood pressure (SAP), respiratory rate (RR) and bispectral index (BIS score) were recorded each 5-min intervals throughout both periods (rest and music). Agitation/sedation state and pain were evaluated by the Richmond-Agitation-Sedation-Scale (RASS) and the Numerical-Rating-Scale (NRS) respectively, before and after each studied periods. Music have not been performed in five patients (5/35 = 14%). Results Music significantly decreased HR (88 ± 15 vs 82 ± 15, P < 0.05), SAP (137 ± 17 vs 128 ± 14, P < 0.05), RR (25 ± 3 vs 22 ± 4, P < 0.05), BIS (94 ± 5 vs 81 ± 10, P < 0.01), RASS (+0.1 ± 0.7 vs –0.7 ± 0.9, P < 0.05) and NRS (4.4 ± 1.7 vs 1.9 ± 1.3, P < 0.01) in both intubated and non-intubated groups whereas no significant change was observed during the rest period. The variations level studied parameters induced by music were comparable for the two groups. Conclusion A single music therapy session was found to be effective for decreasing anxiety and promoting relaxation, as indicated by decreases in heart rate, blood pressure, BIS and respiratory rate over the intervention period in intubated patients during weaning phase.

Pediatric pyeloplasty: Intubated vs nonintubated

Objectives: Despite continued controversy regarding the optimal method of urinary diversion after pyeloplasty, we have treated majority of our patients without any tubes or stents. The present study is done to assess the benefits of performing nonintubated pyeloplasty in children. Materials and Methods: Thirty consecutive cases of pelviureteric junction obstruction were enrolled in the study from 1998 to 2004 (5.6 years). Six patients underwent nephrectomy and were excluded from the study. Twenty-four patients underwent pyeloplasty performed by a single surgeon; 20 were nonintubated pyeloplasties, other four were intubated. Postoperative complications and length of hospital stay were assessed. Results: The intubated group had higher incidence of UTI and persistence of partial mechanical obstruction postoperatively on radio-nucleotide scans. They also had a significantly longer hospital stay as compared with the nonintubated group. Conclusions: Nonintubated pyeloplasty in children is safe, cost effective and associated with fewer complications. A stent and a nephrostomy tube appear to be unnecessary in the routine pyeloplasty.

Outcomes of Patients With Acute Respiratory Failure After Abdominal Surgery Treated With Noninvasive Positive Pressure Ventilation

Little is known about the physiologic and clinical effects of noninvasive positive pressure ventilation (NPPV) in patients who have acute respiratory failure (ARF) after abdominal surgery. We evaluated our clinical experience with the use of NPPV in the treatment of ARF after abdominal surgery. We prospectively evaluated NPPV use during a 2-year period in a medical-surgical ICU of a university hospital. We documented demographic and diagnostic data, gas exchange, and clinical outcomes. We compared patients who were not intubated to those who were intubated after a trial of NPPV. Of 72 patients with ARF after abdominal surgery who were treated with NPPV, 48 patients avoided intubation (67%). Patients in the intubated and nonintubated groups had similar demographic characteristics, and similar American Society of Anesthesiologists physical status and simplified acute physiology score II scores at admission. The intubated group had a significantly lower Pa(O2)/fraction of inspired oxygen (Fi(O2)) ratio (123 +/- 62 mm Hg vs 194 +/- 76 mm Hg, p < 0.01) and more extended bilateral alveolar infiltrates (67% vs 31%, p < 0.01) than the non-intubated group. Within the first NPPV observation period, the Pa(O2)/Fi(O2) increased (+ 36 +/- 29% [+/- SD], p = 0.04) and the respiratory rate decreased (28.2 +/- 3.4 breaths/min vs 23.1 +/- 3.8 breaths/min, p < 0.01) significantly only in the non-intubated group. The non-intubated group had significantly lower length of ICU stay (17.3 +/- 10.9 days vs 34.1 +/- 28.5 days, p < 0.01) and mortality rate (6% vs 29%, p < 0.01). NPPV may be an alternative to conventional ventilation in selected patients with ARF after abdominal surgery who require ventilatory support.

Effects of prenatal alcohol exposure on forepaw digit length and digit ratios in rats

Prenatal exposure to alcohol can cause limb and digit defects. Variations in digit ratios in humans are associated with prenatal testosterone exposure. Since prenatal alcohol can reduce testosterone in rats, the effects of prenatal alcohol were measured on rat digit length. Pregnant Sprague-Dawley rats were intubated with 0 g/kg, 4 g/kg, or 6 g/kg of ethanol from Gestational Day 8 (GD8) to GD20. The 0-g/kg group and a nonintubated group served as controls. At postnatal day 31, forepaw digit lengths were measured and digit ratios calculated. Females had smaller digits on both forepaws and higher digit ratios on the right forepaw than males. Rats exposed to 6 g/kg of ethanol had smaller digits than controls on both forepaws and higher digit ratios than controls on the left forepaw. Rat digit ratios differ between the sexes, and prenatal alcohol exposure affects digit ratios. The results are consistent with a perinatal disruption of testosterone levels by alcohol and/or of testosterone's effects on digit length and ratios. An alternate interpretation is consistent with a retinoic acid-mediated effect of alcohol on digit length and ratios.

Profile of Near-Fatal Asthma in an Inner-City Hospital

Patients requiring mechanical ventilation for asthma are at a high risk of similar attacks and death in the future. Early recognition and treatment with systemic corticosteroids (SC) can influence outcome in near-fatal asthma (NFA). We studied the ability of patients to recognize the severity of asthma, implement SC therapy, and seek timely help from a health-care provider (HCP). Retrospective review of patient histories and medical records. Medical ICU and inpatient medical ward of an inner-city university hospital. Asthma patients requiring mechanical ventilation. Seventy episodes of NFA requiring endotracheal intubation (intubation group [IG]) and 523 hospital admissions with acute asthma (nonintubation group [NIG]) were analyzed over a 4-year period from January 1997 to September 2000. Prior intubation(s), duration of symptoms, steroid dependence, and knowledge and use of inhaled corticosteroids and SC were noted. Indications for mechanical ventilation, ventilatory parameters using permissive hypercapnia, sedation/paralysis, duration, extubation criteria, complications, and outcome were analyzed. Twenty-nine of 70 patients (41.4%) in the IG had at least one prior episode of NFA requiring mechanical ventilation, compared to 123 of 523 patients (23.5%) in the NIG (p < 0.005). Ten of 70 patients (14.3%) in the IG were steroid dependent, compared to 40 of 523 patients (7.6%) in the NIG (p < 0.05). Forty of 70 patients (57%) in the IG compared to 308 of 523 patients (59%) in the NIG reported noncompliance with prescribed inhaled corticosteroids (not significant). All received aerosolized beta-agonist therapy, but only 11 of 70 patients (15.7%) initiated SC therapy in the IG compared to 145 of 523 patients (27.7%) in the NIG (p < 0.05). Even with symptoms > 48 h, 34 of 43 patients (79%) in the IG did not receive SC. Three patients died and four patients acquired myopathy in the IG. History of intubation and steroid dependence were identified as risk factors for future intubation. Our patients with asthma in both the IG and NIG were noncompliant with inhaled corticosteroids. Despite a long duration of symptoms and knowledge of SC, the majority neither implemented self-management with SC nor contacted an HCP. Early initiation of SC (symptoms < 48 h) might have averted intubation in the high-risk group. Most of our patients were extubated successfully in < 48 h. Incidence of barotrauma was very low, probably due to utilization of permissive hypercapnia.

Prehospital intubation in severe thoracic trauma without respiratory insufficiency: a matched-pair analysis based on the Trauma Registry of the German Trauma Society.

On the basis of the data of a multicenter study, the impact of prehospital intubation and ventilation in the therapy of severe thoracic trauma without manifest respiratory insufficiency was analyzed. Data were collected prospectively in the Trauma Registry of the German Trauma Society. In a matched-pair analysis, patients with severe thoracic trauma (Abbreviated Injury Scale score of 4) with and without prehospital intubation were compared. Patients were paired with respect to age, injury severity, and prognosis (according to the TRISS method). From a total of 3,814 patients, two groups (with/without prehospital intubation) of 44 matched patients each with comparable average age (36 vs. 36 years), Injury Severity Score (29 vs. 29), and TRISS (95.2 vs. 95.3) were identified. No patient was unconscious at the scene (all Glasgow Coma Scale scores > or = 8) or presented with severe respiratory insufficiency (all > or = 10 breaths/min). Time between injury and hospital admission was significantly longer (73 minutes; p < 0.05) in the group with prehospital intubation compared with the nonintubated group (47 minutes). Furthermore, fluid requirements in the prehospital period were significantly higher in the intubated patients (3,000 mL vs. 1,000 mL). In the prehospital intubation group, the number of patients with mass transfusion (9 vs. 4) as well as with emergency operations (10 vs. 4) were not significantly different from the nonintubated group. The prehospital intubation group showed a similar incidence of lung failure (17 vs. 14), kidney failure (6 vs. 2), and circulation failure (13 vs. 5). Except for two of the primarily nonintubated patients, all were intubated during their stay in the emergency room or on the intensive care unit. Days of ventilation (median, 7 days) as well as the length of stay on the ICU (median, 11 days) were comparable in both groups. Mortality in the prehospital intubation group was not significantly different between groups (six vs. two deceased). Prognosis with respect to organ failure, treatment time, and mortality is not adversely affected in the German trauma system, if patients with severe thoracic trauma without manifest respiratory insufficiency and without other indications for intubation are not treated with prehospital intubation.

Delivery Room Intubation of Thin Meconium in the Low Risk Newborn: A Clinical Trial † 1058

OBJECTIVE: To determine if intubation of the low-risk newborn with thin meconium affects the incidence of respiratory symptoms.METHODS: Exclusion criterion included moderate or thick meconium, fetal distress, neonatal depression or prematurity. Eligible infants were randomized to either an intubation (Group I) or to a non-intubation group(Group II). The outcome was the presence of respiratory symptoms. This study was approved by the hospital institutional review board. Chi-square, Fisher Exact test,or ANOVA used as appropriate. RESULTS: Patients were studied from May 1994 to June 1997. There were 8,967 births during this period. 7.9% (708/8,967) were delivered through meconium. Thin meconium was noted in 50.3% (356/708) of all births. 24/356 infants with thin meconium were excluded for medical criterion. 169 infants were randomized, 77 infants into Group I and 92 infants into Group II. 163 infants were medically eligible but could not be randomized due to lack of consent, late arrival of the team or obstetrician request. These were retrospectively assigned into intubation(Group IB) and non-intubation (Group IIB) groups. From the intubation groups I and IB, one required supplemental oxygen and was weaned to room air in 7 hours. From the non-intubation groups II and II B, two infants required oxygen, weaning to room air in 11 and 46 hours. Comparing birthweight, gestational age, sex, mode of delivery and 5-minute APGAR there were no significant differences. However, the intubation groups had significantly lower 1-minute APGAR scores. There were no airway morbidities reported in the intubation groups. CONCLUSION: In the infant with thin meconium and an otherwise low risk pregnancy, we were unable to demonstrate a difference in respiratory symptoms with intubation and intra-tracheal suctioning. Table

A selective aggressive approach to the neonate exposed to meconium-stained amniotic fluid

OBJECTIVE: We attempted to determine the effect on meconium aspiration syndrome from a selective approach of neonatal endotracheal intubation in meconium-exposed fetuses. STUDY DESIGN: All pregnancies delivered at the Medical College of Virginia in 1990 were included. Meconium was noted at membrane rupture and qualified as thick, moderate, or thin. Neonates underwent suctioning with a DeLee device at delivery. They were observed without endotracheal intubation if they fulfilled the following criteria: vaginal delivery, gestational age >37 weeks, birth weight >2500 gm, and anticipated Apgar score ≥8 at 1 minute. RESULTS: Of 4289 deliveries, 659 were exposed to meconium and 48% of infants were intubated. Birth weights, gestational ages at delivery, and umbilical arterial pH were similar between intubated and nonintubated neonates. Neonatal intensive care unit admissions were significantly higher in intubated neonates. All 9 neonates diagnosed with meconium aspiration syndrome were intubated at birth. CONCLUSIONS: Utilization of this protocol resulted in reduction of neonatal intubation. Meconium aspiration syndrome did not occur in the nonintubated group. (Am J Obstet Gynecol 1996;175:296-303.)

Recovery of tracheal epithelium following high frequency ventilation at low inspired humidity

We used scanning electron microscopy (SEM) and light microscopy (LM) to study the recovery of tracheal epithelium in newborn lambs damaged by high frequency flow interrupted ventilation (HFFIV) at low inspired humidity (30%). Newborn lambs were mechanically ventilated for 6 h, allowed to recover and subsequently killed at 2, 7 or 14 days. The recovery of the trachea above and below the tip of the endotracheal tube (ETT) was studied at these time periods and compared to a control non-intubated group and a group killed immediately after 6 h of ventilation. Above and below the ETT, SEM and LM revealed deciliation to be greatest 2 days after ventilation. The damaged tracheal mucosa had converted to non-ciliated epidermoid squamous metaplastic cells. Recovery was not complete by 14 days, although the squamous cells had already differentiated into goblet and ciliated columnar epithelial cells. No difference was seen in the rate of recovery of the tracheal mucosa above or below the tip of the ETT.

Comparative Effects of Duodenal and Heal Intubation on Gastric Emptying and Postprandial Antral, Pyloric, and Duodenal Motility

The influence of gastric and small-intestinal intubation on fed patterns of antropyloroduodenal motility and gastric emptying of a solid meal has been investigated in normal volunteers. In 10 subjects a manometric assembly was passed as far as the fourth part of the duodenum; in 8 other subjects the terminal ileum was intubated; and a further 8 subjects were not intubated. The manometric assemblies were similar, apart from their length, and both incorporated a sleeve/side hole assembly located across the pylorus. All subjects ingested a meal of 100 g 99mTc-labelled liver/ground beef. There was no significant difference in either the lag phase or the linear emptying phase of gastric emptying between the control and duodenal intubation groups. The emptying phase of the test meal, but not the lag phase, was slower (p less than 0.01) in the ileal intubation group than in the non-intubated and duodenal intubation groups. There were fewer (p less than 0.01) antral pressure waves in the ileal than in the duodenal intubation group. We conclude that ileal but not duodenal intubation has an important inhibitory influence on gastric emptying and antral motility.

Long‐Term Enteral Feeding of Aged Demented Nursing Home Patients

The terminal phase of dementia is initiated by the inability to swallow. New techniques of enteral alimentation permit more effective, longer intubation. To assess the application of these new techniques to late-stage demented aged patients, all current intubations in a teaching nursing home were reviewed. Of 52 feeding intubations, 26 had been in situ for more than 1 year. A randomly selected comparison group of nonintubated patients was also studied. Weight increased for 48% of the intubated group versus 17% of the nonintubated group (P less than .01). Aspiration pneumonia occurred more often in the intubated group (58%) than in the nonintubated group (17%) (P less than .01). Decubitus ulcers were also more common in the intubated group (21%) than in the nonintubated group (14%). Restraints were used more in the intubated group (71%) than in the nonintubated group (56%). These differences did not reach statistical significance. All of the intubated patients were severely demented, with MMSE scores of zero. Seventy-one percent of the nonintubated group were demented, with MMSE scores of less than 23. Prolongation of the terminal phase of dementia in the aged by tube feeding is now feasible. The implications of this change in the life-span of demented nursing home patients need attention by families, nursing homes, and those who make public health policy.

Priorities in Urinary Diversion Following Pyeloplasty

Debate continues regarding optimal drainage following pyeloplasty. We reviewed the method of drainage in 100 consecutive dismembered pyeloplasties performed from 1983 to 1987. Of the children 20 per cent underwent pyeloplasty with placement of a nephrostomy tube and transanastomotic stent (group 1), while 80 per cent underwent pyeloplasty performed in a nonintubated fashion (group 2). A urethral bladder catheter was used in 20 per cent of the patients in group 1 and in 90 per cent in group 2. While the over-all complication rates and long-term results were similar for both groups, the nature of complications for each group differed. The intubated group had a 10.5 per cent incidence of urinary tract infection, while the nonintubated group demonstrated an 8.6 per cent incidence of postoperative anastomotic leakage. A nephrostomy tube and stent appear to be unnecessary in the routine pyeloplasty and may prolong hospitalization. Selective use of upper tract drainage may help to decrease the incidence of postoperative leakage in specific cases.

Nasogastric intubation After intestinal Resection

Despite some evidence that gastric decompression may be unnecessary after some abdominal operations and in the treatment of paralytic ileus, the use of nasogastric suction after extensive abdominal operations, particularly intestinal resection, remains a subject of some debate. In a randomized prospective trial, 52 patients with suture lines constructed in the gastrointestinal tract received no postoperative nasogastric drainage unless acute gastric dilation or copious vomiting developed postoperatively, while 45 similar patients were allocated to receive routine postoperative nasogastric aspiration. Only 12 patients in the nonintubated group required subsequent insertion of a nasogastric tube, while in the remaining 40, nasogastric drainage was avoided completely. Postoperative loss of fluid from the intestine was significantly greater in the patients undergoing routine nasogastric drainage, although the requirement for intravenous fluid therapy was the same in both groups. There were no significant differences between the two groups in the incidence of postoperative complications. Routine nasogastric aspiration after gastric or intestinal resection does not confer significant advantages to outweigh its discomfort and potential morbidity for patients and should be replaced by selective intubation when required postoperatively.