Monitoring respiration and oxygen saturation in patients during the first night after elective bariatric surgery: A cohort study

Background: Obstructive sleep apnoea and obese hypoventilation is not uncommon in patients with obesity. Residuals effect from surgery/anaesthesia and opioid analgesics may worsen respiration during the first nights after bariatric surgery. The aim of this observational study was to monitor respiration on the first postoperative night following elective bariatric surgery. Methods: This observational study aimed to determine the incidence and severity of hypo/apnoea in low risk obsess patients undergoing elective bariatric surgery in general anesthaesia. Patients with known or suspected sleep respiratory disturbances was not included. ESS was scored prior to surgery. Oxygen desaturation was analyzed by continuous respiratory monitoring. Mean oxygen saturation (SpO2), nadir SPo2, apnoea/hypopnea index and oxygen desaturation index was assess by standard tools. Results: 45 patients were monitored with portable polygraphy equipment (Embletta, ResMed) during the first postoperative night at the general ward following elective laparoscopic bariatric surgery. The prop ESS was 0-5 in 22, 6-10 in 14 and 11-16 in 6 of the patients studied (missing data 3). Mean SpO2 was 93%; 10 patients had a mean SpO2 of less than 92% and 4 of less than 90%. The lowest mean SpO2 was 87%. There were 16 patients with a nadir SpO2 of less than 85%, lowest nadir SpO2 being 63%. An Apnoea Hypo/apnoea Index (AHI) > 5 was found in 2 patients only (AHI 10 and 6), and an Oxygen Desaturation index (ODI) > 5 was found in 3 patients (24, 10 and 6, respectively). 3 patients had more prolonged (> 30 seconds) apnoea with nadir SpO2 81%, 83% and 86%. ESS score and type of surgery did not impact on respiration/oxygenation during the observation period. Conclusions: A low mean SpO2 and episodes of desaturation were not uncommon during the first postoperative night following elective bariatric surgery in patients without history of night time breathing disturbance. AHI and/or ODI of more than 5 were only rarely seen. Night-time respiration monitoring provided seemingly sparse additional information. Further studies are need to assess risk factors and potential impact of the desaturation episodes that occurs during sleep.

: This observational study aimed to determine the incidence and Methods severity of hypo/apnoea in low risk obsess patients undergoing elective bariatric surgery in general anesthaesia. Patients with known or suspected sleep respiratory disturbances was not included. ESS was scored prior to surgery. Oxygen desaturation was analyzed by continuous respiratory monitoring. Mean oxygen saturation (SpO2), nadir SPo2, apnoea/hypopnea index and oxygen desaturation index was assess by standard tools.
: 45 patients were monitored with portable polygraphy equipment Results (Embletta, ResMed) during the first postoperative night at the general ward following elective laparoscopic bariatric surgery. The prop ESS was 0-5 in 22, 6-10 in 14 and 11-16 in 6 of the patients studied (missing data 3). Mean SpO2 was 93%; 10 patients had a mean SpO2 of less than 92% and 4 of less than 90%. The lowest mean SpO2 was 87%. There were 16 patients with a nadir SpO2 of less than 85%, lowest nadir SpO2 being 63%. An Apnoea Hypo/apnoea Index (AHI) > 5 was found in 2 patients only (AHI 10 and 6), and an Oxygen Desaturation index (ODI) > 5 was found in 3 patients (24, 10 and 6, respectively). 3 patients had more prolonged (> 30 seconds) apnoea with nadir SpO2 81%, 83% and 86%. ESS score and type of surgery did not impact on respiration/oxygenation during the observation period.
: A low mean SpO2 and episodes of desaturation were not Conclusions uncommon during the first postoperative night following elective bariatric surgery in patients without history of night time breathing disturbance. AHI and/or ODI of more than 5 were only rarely seen. Night-time respiration monitoring provided seemingly sparse additional information. Further studies are need to assess risk factors and potential impact of the desaturation report report report report monitoring provided seemingly sparse additional information. Further studies are need to assess risk factors and potential impact of the desaturation episodes that occurs during sleep.

Introduction
Obesity is on the increase in the western world and is associated with the development of several diseases. It is a major risk factor for cardiovascular disease and diabetes, two of the leading causes of death globally. Many efforts have been made in trying to treat the condition (http://www.who.int/mediacentre/factsheets/ fs311/en/). The adverse effects on pulmonary function are also well documented 1 . With an increasing BMI, the risk for chronic daytime hypoventilation escalates, characterized by an arterial carbon dioxide pressure (PCO 2 ) exceeding 45 mmHg 1 . Complications include atelectasis, hypoxemia, pulmonary embolism and subsequent acute ventilation failure, and may evolve during perioperative and postoperative phases 1 . Adverse effects are not concentrated to daytime; obesity is the most frequent predisposing factor of obstructive sleep apnoea syndrome (OSAS) 1 . Changes in the breathing pattern and pulmonary function may indeed cause compromise of oxygenation and increased risk for oxygen desaturation.
The early postoperative period with residual anaesthetic and analgesic effects may put an obese patient that has just undergone a laparoscopic bariatric procedure at risk for respiratory compromise. It has been debated whether early care should be carried out in a high dependency ward or could be safely done in a general ward.
The present study aimed to monitor first postoperative night respiration, breathing patterns and oxygenation with sleep breathing equipment, in patients having undergone elective laparoscopic bariatric surgery. We wanted to assess whether we could define risk factors associated with hypo/apnoea and desaturation episodes.

Methods
This is an explorative cohort study; the study protocol was approved by the ethical committee at Karolinska Institutet [Dnr 2015/118 -31/1 La]. Patients were included in the study after having provided verbal and written informed consent. Patients with known sleep related respiratory disturbances, diagnosed or suspected obstructive sleep apnoea was excluded.
Each patient filled out a questionnaire to determine if there was a suspicion of OSAS preoperatively; the questionnaire included the Epworth Sleepiness Scale (ESS). Patients were monitored after surgery from 10 pm during the first postoperative night, until 6 am the next morning, with a portable OSAS breathing pattern monitor Embletta (ResMed). The registration included information about airflow from a nasal cannula, thoracic respiratory movements by an elastic band around the thorax and percutaneous O2 saturation and heart rate from a pulse oximeter.
Apnoea was classified in accordance to the American Academy of Sleep Medicine (AASM) as a drop in the peak signal excursion by ≥ 90% of pre-event baseline air flow signal. The duration of the ≥90% drop in sensor signal must be ≥10 seconds. Hypopnea was classified by as a drop in the peak signal excursion by ≥ 30% of pre-event baseline. The duration of the ≥ 30% drop in signal excursions must be ≥ 10 seconds (Berry et al., 2012) 2 .
All patients had anaesthesia and postoperative pain management in accordance to the routines of the department. All patients received premedication with 2 tablets of slow release 655mg paracetamol and 10 mg slow release oral oxycodone prior to surgery. Patients were preoxygenated by FiO2 1.0 and by CPAP of 6 cm H20 in the anaesthetic machine Aisys (GE Healthcare). Anaesthesia was induced with remifentanil target control infusion (TCI), set at a target of 6.0 ng/ml. After 90 seconds the patient was put to sleep with a bolus injection of propofol 2-3 mg/kg. When the patient got apnoeic the ventilation mode was changed to pressure control ventilation-volume guaranteed (PCV-VG), and the patient received neuromuscular blocker rocuronium, followed by endotracheal intubation. Anaesthesia was maintained with sevoflurane and remifentanil titrated to clinical signs of adequate anaesthesia, and with a BIS (Medtronic, Covidien BIS LoC 2 Channel) value between 25 and 50. All patients received postoperative nausea and vomiting (PONV) prophylaxis with betamethasone, ondansetron and droperidol. 10 -15 mg of morphine was administered at the beginning of surgery, as a start dose for the postoperative pain relief regime. The patient had laparoscopic surgery, gastric bypass or sleeve gastrectomy, as decided by the surgeon. Postoperative care was provided in accordance to routines; fentanyl 25-50 micrograms was used as rescue analgesia and a further 1-5 mg morphine administered as needed. Postoperative respiratory care included oxygen supplementation to satisfactory saturation and once per hour blowing in a T-piece with a one-way valve mouthpiece (Intersurgical). Patients were transferred to the general ward when fully awake and with stable vital signs for 30 minutes. No intervention was done apart from the night breathing monitoring.

Statistics
All data is presented as mean and standard deviation. The breathing data was evaluated in accordance to standard assessment, • the AHI was calculated as hypo and/or apnoea/hour • the ODI was calculated as oxygen saturation decrease of > 4 for 30 seconds/hour

Results
There were 52 patients initially included in the study, but 6 were excluded as they had a diagnosis of OSAS and 1 further patient was

Amendments from Version 1
The paper has been amended/revised extensively. The referees provided important and most adequate comments. The Method section has been revised, describing exclusion criteria, method for assessing breathing abnormalities. The Result section has been revised taking the preoperative ESS score into account and also providing additional analysis based on the ESS. Table 3, presenting the findings of respiration and oxygenation in relation to preoperative ESS score, has been added to the Results section. The Discussion is amended and the potential risks are further addressed. The Conclusion is revised stressing the need for further studies.

See referee reports
REVISED excluded as the procedure became merely a diagnostic laparoscopy. Forty-five patients were included in the study.
The mean age for the 45 patients studied was 39 (ranging from 19 -68 years), and the mean BMI was 37 (ranging from 32 -53). The preoperative ESS scale screening showed that 22 of our patients had an ESS 0-5, 14 had a score of 6 -10 and 6 patients had ESS of more than 10. Highest ESS was 16 in one 26 year old female patient with a BMI of 41.
Surgery and anaesthesia was uneventful, mean duration of surgery was 54 minutes (ranging from 27 -97 minutes) and mean duration of anaesthesia was 108 (ranging from 65 -217 minutes).
Most patients (n=35), had a sleeve gastrectomy and 10 had a gastric bypass procedure.
Mean time between end of anaesthesia and start of the polygraphy was 611 ± 122 minutes and mean duration of polygraphy monitoring was 463 ± 51 minutes.
Mean saturation (SpO2) during the polygraphy was 93% (ranging from 87 -97). There were 10 patients with a mean SpO2 of less than 92% and 4 with mean of less than 90, with the lowest mean SpO2 being 87%. There were 16 patients with a nadir SpO2 of less than 85%, lowest nadir SpO2 being 63%.
Changes in respiration and oxygenation did not differ between the 3 ranges of ESS scores, see Table 1.
We could further not see any difference apart from a shorter duration of anaesthesia between the surgical procedures (See Table 2 and

Discussion
We found, somewhat surprisingly, only very minor respiratory disturbances in the cohort of patients having undergone elective bariatric surgery. No patients had a hypopnea index above thirty -only 2 had an AHI above 5. A majority of patients had low oxygen saturation of around 93%, and short episodes of saturation below 85% were not uncommon. Thus, the main finding is mild hypoxia and episodes of desaturation, but hypo/apnoea monitoring does not provide much additional information. Signs of more pronounced airway compromise causing hypo/apnoea were only rarely seen.  Low mean saturation and desaturation episodes were however not uncommon. The preoperative ESS score did not help predict differences in oxygenation. All our patients had no further complications.
The risk for postoperative hypoxia has been known for long. Jones et al. published a review in 1990 addressing the risk for low oxygenation during the early postoperative period, and its multifactorial etiology 3 . Low saturation and mild hypoxia has also been reported in previous studies and the risk in bariatric surgery must be acknowledged 4 .
We are not aware of any previous study explicitly monitoring hypo/ apnoea during the first postoperative night after bariatric surgery Zaremba et al. studied polysomnography in patients during the early postoperative course, while patients were still in the PACU 5 . They found that 64% of the 33 patients with complete postoperative polysomnography data had signs of sleep-disordered breathing with an AHI greater than 5/h early after recovery from anaesthesia. The respiratory response to hypoxia and hypercapnia caused by airway obstruction is compromised following major surgery 6 . Chung et al. studied respiration during the preoperative stage and the first, second and third postoperative nights 7 . Our results are in line with these results, first night after surgery. Female patients and patients with no or mildly compromised nocturnal breathing showed only minor changes during the first postoperative night in Chung et al's study as well as in ours. Age, preoperative respiration disturbance and smoking were found to be risk factors for hypo/apnoea 8 .
We used a standard portable breath and saturation monitor, the Embletta system. The portable systems have been shown to be accurate tool for assessment of sleep apnoea 9 . We did not include EEG monitoring and we did not attach the abdominal movement tracing. Merely the nasal flow, thoracic and saturation was recorded. AHI scores were > 5 for mild, > 15 for intermediate and > 30 for significant airway compromise, or "sleep apnoea". We assessed SpO2 < 94 as mild hypoxia. The monitoring was initiated during the first evening, an average of ten hours after the end of anaesthesia. It should also be acknowledged that the mean BMI in our cohort was 37, versus s mean BMI of 44 in the US studies 5 .
Nocturnal oxygen desaturations are not uncommon during the first postoperative night, also in patients undergoing other surgical procedures. Shirmakana et al. found that oxygen desaturation was frequent in patients having undergone breast surgery 10 . Bowdle looked at a mixed group of ambulatory surgery patients and also found an increase in hypo/apnoea and desaturation in the first night after surgery 11 .
There are recent guidelines from the US association for sleep apnoea 12 . Germany has also addressed the importance of adequate perioperative care of obese patients at risk for sleep apnoea 13 . The preoperative evaluation by registration of hypo/apnoea is strongly recommended, however advice given around the early postoperative period is sparse. The ESS should of course also be used for screening 14 . The waste majority of our patient had a "normal" ESS, and group of patients with ab ESS above 10 was small and did not provide any firm additional information.
To put our findings into perspective, our patients had a mean BMI of merely 37. Additionally, we had a majority (39 out of 45) of female patients, and BMI-associated AHI abnormalities are more commonly seen in males 15 . The physical status, muscle strength and breathing capacity was not assessed or screened preoperatively. We did not use the abdominal movement monitoring band to avoid additional abdominal pain. All patients had multi-modal analgesia and opioids were administered as restrictively as possible to maintain adequate pain control. There is an obvious need for further studies looking for risk factors and also better determine potential risks associated to the desaturation episodes that are not uncommonly noticed.
In conclusion, we have found that elective "low risk" obese patients that have had uncomplicated laparoscopic bariatric surgery, have low saturation during the first postoperative night and may experience episodes of oxygen saturation of less than 85%, but hypo/apnoea is rare and monitoring of obstruction seems not to be of major value. The potential clinical impact of the mild hypoxia and short episodes of desaturations requires further study.
We have followed the STROBE guidelines. Author contributions JJ and SF were in charge of the study design protocol and processed applications. LW and SF had major roles in patient recruitment and collection of study data. All authors contributed equally to data analysis and manuscript preparation.

Competing interests
No competing interests were disclosed.

Grant information
Financial support was obtained from the research department at TioHundra Norrtälje.
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. This observational study sought to determine the incidence and severity of hypopnea or apnea on the first night after surgery, in patients who did not have a diagnosis of OSA undergoing elective laparoscopic bariatric surgery. They monitored 45 patients with a portable polygraphy during the first postoperative night. They report that oxygen desaturation and short episodes of desaturation occurred in some patients, but there were few patients with an AHI or ODI greater than five. They conclude that respiratory monitoring for low risk patients on the general ward provides little additional information.

Open Peer Review
Although this is an interesting study, this manuscript can be improved by providing more details in the Methods and Results. In the Introduction, the authors state that they wished to assess risk factors associated with hypopnea or apnea and desaturation. However, the presence of co-morbidities is not mentioned and there is no analysis of the risk factors that were associated with the patients that experienced desaturation/apnea/hypopnea. The questionnaire that was used to determine the risk of OSAS is indicated to be Epworth Sleepiness scale. The validation of this scale to screen for sleep apnea is relatively modest. It is not clear whether patients with obstructive sleep apnea are excluded from the study after use of questionnaire. Is sleep apnea one of the exclusion criteria of the study? The results of the questionnaire to determine risk of OSAS should be reported, it is unclear whether the patients who participated in the study were at low risk or may be at risk or have undiagnosed OSAS.
The definition of apnea/hypopnea should be stated. The explanation for the sample size should be added. The number of patients receiving oxygen on the ward should be added, especially as it is mentioned in the methods section that patients received oxygen supplementation postoperatively. In Tables 1 and 2, the authors compare gastric bypass vs. sleeve gastrectomy, however, the authors have not stated whether they expected a difference between these 2 groups. There are some grammatical errors in the manuscript.

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Partly

Are the conclusions drawn adequately supported by the results? Partly
No competing interests were disclosed.

Competing Interests:
We have read this submission. We believe that we have an appropriate level of expertise to 2 We have read this submission. We believe that we have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however we have significant reservations, as outlined above. We have analysed the results also taking the ESS into account, but we failed to see any clear association between ESS and respiration/saturation see Table 3  Table 3 ESS We did not know whether surgical procedure had an impact and therefore tested also for surgical technique.
We hope that these responses is in line with your expectations, an update manuscript is on the way. Hypoxemia in the immediate postoperative period is a well-known and feared complication after bariatric surgery. Risk factors for hypoxemia include patient age (likely because of the increase in closing volume with aging), loss of lung volume during induction and maintenance of general anesthesia, alterations in respiratory drive related to general anaesthetics and use of opiates, as well as the the presence of obstructive sleep apnea.
The exact incidence of hypoxemia and hypercarbia after bariatric surgery has not been well studied, but several studies in relatively small numbers of patients have documented worrisome levels of hypoxemia.
Wickerts and co-workers have studied 45 good risk bariatric surgery patients. The mean age of the cohort was 39, the mean BMI was 37, and patients with known Obstructive Sleep Apnea were excluded. They all underwent uncomplicated and brief laparoscopic foregut procedures for obesity and were studied for 8 hours with pulse oximetry during the first postoperative night. Study findings included 22% with a mean SaO2 < 92%, 9% with mean SaO < 90% (lowest mean SaO was 87%), and 36% with nadir SaO 1,2 2,3 4 for 8 hours with pulse oximetry during the first postoperative night. Study findings included 22% with a mean SaO2 < 92%, 9% with mean SaO < 90% (lowest mean SaO was 87%), and 36% with nadir SaO < 85% (lowest nadir SaO was 63%). In addition, sleep disordered breathing events were documented in several patients. No clinical events related to hypoxia or sleep disordered breathing are recorded.
These findings are important as because they do document a small, but significant risk of potentially dangerous postoperative hypoxia even in low risk patients. Recently published guidelines recommend continuous monitoring with pulse oximetry in the early postoperative period. Guidelines for perioperative management of low risk patients as studied here are lacking.
Additional studies of the epidemiology of perioperative hypoxemia and hypercarbia complicating bariatric surgery are needed in order to begin to identify potentially modifiable patient-specific risk factors and interventions such is inspiratory muscle training which might reduce the risks of this potentially dangerous problem. The authors are encouraged to continue this survey and attempt to relate hypoxemia to risk factors. In addition, continued studies to include EKG monitoring with pulse oximetry might add to the clinical significance of these findings. The authors comment that the observed episodes of desaturation were brief, but they do not share any data regarding the duration of desaturations. I would be concerned about the possible clinical impact of 30 seconds of apnea in an unmonitored setting on the initial post operative night following bariatric surgery In this era of cost constraints and limited patient access to bariatric surgery, cost reduction strategies designed to encourage same day bariatric surgery procedures are popular. Improved ability to recognize and address risk factors for postoperative hypoxemia should be an important component of the selection process for accelerated care. Thank you for your important and constructive comments on our paper. Bariatric surgery do increase and enhanced recovery pathways are not uncommonly used. The aim of our study was to assess respiration/oxygenation in "low risk patients" undergoing elective laparoscopic "low risk bariatric surgery" and my expectation was indeed that we should observe more. Still agree even the observations found should be interpreted in a context; avoiding patient risk.

Are sufficient details of methods and analysis provided to allow replication by others? Yes
We did not monitor duration of desaturation unfortunately but the hypopnea time was registered and was in two patients prolonged. The portable device used did not log duration of desaturation. We did not monitor ECG and/or sign of myocardial ischemia. There were however no complaints of chest pain or any other signs/symptoms alerting us to suspect myocardial or other ischemic events.
We were in the group of patients not able to identify explicit risk factors.
The literature is extensive and we are more than happy to add suggested papers.
We are planning to continue our efforts looking a´t the first nigh respiratory events and will in these studies add ECG.
I hope this response can be seen as acceptable and further suggestions for future studies are most welcome.

Best regards
Best regards Jan Jakobsson, on behalf of the authors No competing interests were disclosed. Competing Interests: There are of course limitations and we have tried to address some of them but we did not compile a specific strength and limitation section, We studied low risk patients undergoing elective bariatric surgery in one center. All patients followed the same clinical pathway and was cared for by a limited number of physicians and nurses, thus perioperative care was standardized. We used a portable respiratory monitoring equipment so called Embletta. There are limitations with this technique and a more advanced polysomography device would have been an option. We did not monitor ECG, and no other specific monitoring of potential ischemic episodes. We did not measure the carbon dioxide tension. A continuous transcutaneous carbon dioxide monitor could have been of interest.

Discuss this Article
I hope these brief lines can at least to some extent respond to your queries, Best regards Jan Jakobsson on behalf of the authors. In the Statistics section it is better to mention the software that you used for analysis. In the Discussion section it is recommended that explain Limitations & strengths of your study.

Best Wishes
No competing interests were disclosed. Competing Interests: