Muscular Tissue Oxygen Saturation and Posthysterectomy Nausea and Vomiting: The iMODIPONV Randomized Controlled Trial

The intervention guided by Muscular Oxygenation to Decrease the Incidence of PostOperative Nausea and Vomiting (iMODIPONV) was a multicenter, pragmatic, superiority, patient- and assessor-blinded randomized controlled (1:1 ratio) trial comparing the effects of intraoperative muscular tissue oxygen saturation–guided care and usual care on postoperative nausea and vomiting in patients having laparoscopic hysterectomy. The trial was designed and overseen by a steering committee (Supplemental Digital Content, https://links.lww.com/ALN/C353), organized at Yale University School of Medicine (New Haven, Connecticut), supported by institutional and departmental sources of the participating institutions, and conducted at six teaching hospitals located in four different cities in China after approval by the institutional review board at each participating hospital. This trial (identifier NCT03641625) was registered at ClinicalTrials.gov by the principal investigator, Lingzhong Meng, on August 22, 2018. The trial was conducted in accordance with the principles of the Declaration of Helsinki and the original protocol. The authors assume responsibility for the accuracy and completeness of the data and analyses, as well as the fidelity of the trial and this report.

We studied patients who had three or more risk factors for postoperative nausea and vomiting, i.e., female sex, nonsmoking status, and expected opioid analgesia use.^12,13  The inclusion criteria were age 18 to 65 yr, nonsmoker status, American Society of Anesthesiologists (Schaumberg, Illinois) Physical Status classification of I to III, and an elective laparoscopic procedure involving hysterectomy. We excluded those patients who were scheduled for vaginal or open hysterectomy, urgent or emergent surgery, or a procedure involving bowel resection. We also excluded patients who had major systemic comorbidities or had undergone chemotherapy or radiotherapy within 3 months before surgery.

The participants were identified and consented on the day before surgery. Both verbal and written informed consent was obtained. On the day of surgery (approximately 30 to 60 min preoperatively), the participants were randomly assigned in a 1:1 ratio based on block randomization (size = 4), using a statistical package (version 9.3; SAS Institute Inc., USA), to the muscular tissue oxygen saturation–guided care group or the usual care group. Each study sequence number (from 1 to 800) and its corresponding group assignment were printed on paper, which was then concealed in a sequentially numbered envelope. The sequence number of the envelope was the same as the study sequence number printed on the paper. All envelopes were kept at the study headquarters and distributed, sequentially and in multiples of four, to each study site based on the speed of patient recruitment. The study envelopes were secured in a dedicated container and managed by the study coordinator at each participating hospital. An envelope was allocated to a participant sequentially.

The study coordinator at each participating hospital and the anesthesia team taking care of the patients were not blinded to group assignment. Other care providers, patients, and outcome assessors were blinded to group assignment. Outcome assessors were not allowed to participate in other aspects of the study and recorded outcome data independently. The study coordinators and investigators participating in intraoperative care were not allowed to participate in outcome assessment. Research laptops used for data capturing were password-secured. All data were first recorded on a predesigned paper case report form and later transferred to a web-based REDCap database (https://projectredcap.org/). Members of the quality committee conducted regular study site visits to reinforce the execution of the study protocol, monitor the accuracy of research data, and ensure the adherence to timely and accurate data recording and transfer. An independent three-member data safety monitoring board was established to monitor patient safety for the trial.

In addition to routine monitoring, including electrocardiography, pulse oximetry, and noninvasive blood pressure, patients from both groups were also monitored using a tissue oximeter (FORE-SIGHT Elite, CAS Medical Systems Inc., acquired by Edwards Lifesciences, USA) and a hemodynamic monitor (LiDCOrapid; LiDCO, United Kingdom). One biophotonic sensor was placed on the left paraspinal muscle (perpendicularly aligned to the spine at the L2–3 level) to monitor the flank muscular tissue oxygen saturation, while another biophotonic sensor was placed on the brachioradialis muscle of the left forearm (approximately two fingers below the antecubital crease) to monitor the forearm muscular tissue oxygen saturation. The finger cuff of the LiDCOrapid monitor was placed on the left hand to monitor systemic hemodynamics, while the cuff used for noninvasive blood pressure measurement as part of the routine care was placed on the right upper arm. The real-time muscular tissue oxygen saturation and hemodynamic data were captured by a research laptop at a frequency of 0.5 Hz. The monitoring and data recording were started before anesthesia induction and stopped immediately before moving the patient from the operating table to the transporting bed.

The anesthetic care was standardized. After facemask preoxygenation, anesthesia was induced using lidocaine, sufentanil, and propofol. All patients were endotracheally intubated and mechanically ventilated with a tidal volume of approximately 6 ml/kg and respiratory rate of 10 to 16 breaths/min to maintain the end-tidal carbon dioxide at 30 to 40 mmHg. The inspired fraction of oxygen was 50%. Anesthesia was maintained using propofol and remifentanil infusion. Dexamethasone and 5-HT3 antagonists were used for postoperative nausea and vomiting prophylaxis. Sufentanil was the drug of choice for intraoperative analgesia.

After patient arrival in the operating room, we first performed baseline measurements, including flank and forearm muscular tissue oxygen saturation, cardiac output, stroke volume, heart rate, systemic vascular resistance, and blood pressure, before anesthesia induction and with the patient awake, calm, and breathing room air. Patients received light sedation if they appeared anxious. Anesthesia was induced after baseline measurements. In the usual care group, the screens of both the tissue oximeter and hemodynamic monitor were covered by an opaque cloth, and the patients were managed per the usual care. In the muscular tissue oxygen saturation–guided care group, both monitors were open to anesthesiologists, and the patients were managed per the muscular tissue oxygen saturation–guided care (fig. 1).

The goal of the muscular tissue oxygen saturation–guided care was to maintain the flank and forearm muscular tissue oxygen saturation equal to or greater than the baseline measurement or 70%, whichever was higher, based on the results of a previous observational study.⁹  Interventions were initiated whenever muscular tissue oxygen saturation was lower than the goal for 60 s at either the flank or forearm location. The first step was to check cardiac output. If cardiac output was lower than the baseline level, stroke volume and heart rate were assessed to determine the cause of cardiac output reduction (cardiac output = stroke volume × heart rate). If stroke volume was lower than baseline, preload augmentation, myocardial contractility enhancement, and/or afterload reduction were considered. If heart rate decrement was the cause, measures to increase heart rate were instituted. The second step was to check systemic vascular resistance if the cardiac output was greater than or equal to baseline. Measures to decrease systemic vascular resistance were considered if the systemic vascular resistance was higher than 80% of baseline. We recommended nicardipine if a vasodilatory agent was deemed necessary (based on unpublished data). The third step was to check blood pressure if both cardiac output and systemic vascular resistance were acceptable while muscular tissue oxygen saturation remained low. Measures to increase blood pressure were considered if it was lower than 80% of baseline.¹⁴  Because blood pressure is proportional to cardiac output and systemic vascular resistance, the differential diagnosis and therapeutic options for hypotension revolved around the first and second steps aforementioned. The following interventions dealt with arterial blood oxygen content, muscular tissue metabolic activity, and other potential causes. We maintained pulse oxygen saturation at 95% or above and hemoglobin at 7 to 9 g/dl depending on the patient’s comorbidities. The intraabdominal pressure was decreased if it was deemed unnecessarily high. Muscle relaxants could be considered when all other measures failed.

The intervention protocol was thoroughly explained to all investigators. Two formal mandatory trainings, 8 h each, were conducted in August 2018 and January 2019. Members of the quality committee (Supplemental Digital Content, https://links.lww.com/ALN/C353) visited all participating hospitals during the first 3 months of the study to resolve outstanding issues and reinforce protocol execution. Remote research meetings were conducted weekly to discuss various issues and coordinate research efforts.

The effectiveness of the intervention protocol in maintaining muscular tissue oxygen saturation in the predetermined ranges was assessed using the below-goal area under the curve, which is the accumulative product of the difference between the goal and the actual muscular tissue oxygen saturation times the duration over which muscular tissue oxygen saturation was below the goal. A smaller area under the curve indicated better intervention effectiveness and better adherence to the intervention protocol, whereas a large area under the curve indicated the opposite (fig. S1A and S1B in Supplemental Digital Content, https://links.lww.com/ALN/C353). Similarly, the degree of hemodynamic changes below the baseline measurement was assessed using the below-baseline area under the curve. A smaller area under the curve indicated a smaller hemodynamic deviation from the baseline and vice versa.

The primary outcome was the incidence of 24-h postoperative nausea and vomiting, defined as the development of nausea, retching, or vomiting within 24 h after surgery.^12,15  The secondary outcomes were the incidence of postoperative nausea and vomiting and the severity of nausea and pain at very early (0 to 2 h), early (2 to 6 h), and late (6 to 24 h) postoperative stages. Patients used an 11-grade numerical rating scale ranging from 0 (none) to 10 (the worst) to rate the severity of their nausea and pain. Patients were considered to experience moderate to severe nausea or pain if the numerical rating scale score was greater than or equal to 5. Other secondary outcomes included (1) the quality of recovery over the first 24 h after surgery as measured by the 15-item quality-of-recovery scale¹⁶ ; (2) the quality of the first night’s sleep after surgery; (3) the time from the end of surgery to the first out-of-bed ambulation; (4) the time from the end of surgery to the first tolerated meal; (5) the length of the patient’s hospital stay; (6) admission to the intensive care unit; (7) hospital readmission; and (8) 30-day morbidity and mortality.

We assumed that muscular tissue oxygen saturation–guided care could reduce the incidence of 24-h postoperative nausea and vomiting from 50%^2,13,17  to 40% in patients having laparoscopic hysterectomy. With two-sided significance and power set at 0.05 and 80%, respectively, the sample size required to detect the projected reduction is 388 patients/group. We planned to enroll 400 patients in each group, assuming a drop-out rate of approximately 3%.

Descriptive statistics were calculated for demographic, baseline, and perioperative characteristics and outcomes. Categorical variables are presented as frequencies and percentages, whereas continuous variables are presented as means and standard deviations or medians and interquartile ranges depending on the data distribution assessed using histograms and Q–Q plots. Absolute standardized differences between the intervention and control groups were calculated using the method described by Yang and Dalton.¹⁸ 

The primary analyses were performed based on the modified intention-to-treat population, which included all patients who had undergone randomization and surgery. The primary outcome was evaluated by the chi-square test, and other binary variables were evaluated by the chi-square test or Fisher’s exact test as appropriate. The effect between two groups was quantified by the risk ratio and 95% CI. CIs for the difference of two proportions were calculated using Newcombe’s method with continuity correction.¹⁹  The two-sample t test or nonparametric Wilcoxon rank-sum test was used to compare continuous variables. CIs for median differences were calculated using Hodges–Lehmann estimates.²⁰ 

A series of sensitivity analyses were performed for the primary outcome. First, a per-protocol analysis was performed based on the modified intention-to-treat population with the exclusion of those patients in whom the study protocol was not followed, including randomization errors, ineligible surgery, nonadherence to the intervention protocol, monitor malfunctioning, and postoperative intubation of the patient. Second, a multiple logistic regression model for the incidence of 24-h postoperative nausea and vomiting was fit in which hospital was adjusted for to assess the impact of between-hospital heterogeneity on risk estimates. Third, another logistic regression model was fit in which hospital, postoperative nausea, and vomiting risk factors and variables having between-group imbalance (P < 0.100) were adjusted for as covariates. The odds ratios from logistic regression analyses were converted into risk ratios using the method of Zhang and Yu.²¹ 

Subgroup analyses of the primary outcome were performed by the following prespecified criteria: participating hospital, age ranges (18 to 39, 40 to 49, 50 to 65 yr), body mass index (less than 25 vs. 25 or more), education level (lower than college vs. college or higher), diagnosis (benign vs. malignant), American Society of Anesthesiologists Physical Status classification (I vs. II to III), baseline hemoglobin level (less than median vs. median or higher), baseline creatinine level (less than median vs. median or higher), history of hypertension, postoperative nausea and vomiting and motion sickness, intraoperative drugs (propofol, remifentanil, sufentanil, and muscle relaxant), anesthesia time (less than 120 min vs. 120 min or more), crystalloid administered (less than median vs. median or higher), estimated blood loss (less than median vs. median or higher), and urine output (less than median vs. median or higher). The heterogeneity of effects across subgroups was assessed by testing the significance of the treatment-by-group interaction in the multivariable logistic regression models of the postoperative nausea and vomiting incidence.

The statistical analyses were performed with R (version 3.5.2) packages including arsenal,²²  stddiff,²³  fmsb,²⁴  and sjstats.²⁵  A two-sided P value of less than 0.050 was considered statistically significant for the primary outcome. The Holm–Bonferroni method²⁶  was applied to adjust for multiple testing for secondary outcomes, subgroup analyses, and sensitivity analysis of the primary outcome.

From September 2018 to June 2019, a total of 1,049 patients scheduled for an elective procedure involving hysterectomy were evaluated, with 800 patients randomly assigned to receive either muscular tissue oxygen saturation–guided care (n = 400) or usual care (n = 400). The treatment groups had reasonably balanced baseline characteristics (table 1) and balanced baseline tissue oxygenation and hemodynamic measurements (table 2). Because of the cancellation of surgery after randomization for one patient from the usual care group, the modified intention-to-treat population had a total of 799 patients, with 400 in the muscular tissue oxygen saturation–guided care group and 399 in the usual care group (fig. 2).

Muscular tissue oxygen saturation–guided care yielded a lower below-goal flank muscular tissue oxygen saturation area under the curve (50 vs. 140% · min; P < 0.001; fig. S1C and S1D in Supplemental Digital Content, https://links.lww.com/ALN/C353) and forearm muscular tissue oxygen saturation area under the curve (53 vs. 245% · min; P < 0.001; fig. S1E and S1F in Supplemental Digital Content, https://links.lww.com/ALN/C353) compared with usual care; it also yielded a lower below-baseline heart rate area under the curve (849 vs. 1,180 beats; P < 0.001), received more crystalloids (1,600 vs. 1,250 ml; P < 0.001), and produced more urine (300 vs. 250 ml; P < 0.001) compared with usual care; these between-group differences remained significant after adjustment of multiple testing (table 2). Patients in the muscular tissue oxygen saturation–guided care group had lower below-baseline cardiac index area under the curve (71 vs. 91 ml/m²; P = 0.004), systolic blood pressure area under the curve (1,617 vs. 1,976 mmHg · min; P = 0.014), and diastolic blood pressure area under the curve (1,086 vs. 1,376 mmHg · min; P = 0.015); however, these between-group differences were not significant after adjustment for multiple testing (table 2).

In total, 127 patients (32%, 127 of 400) in the muscular tissue oxygen saturation–guided care group and 142 patients (36%, 142 of 399) in the usual care group had 24-h postoperative nausea and vomiting, which was not statistically significantly different (risk ratio, 0.89; 95% CI, 0.73 to 1.08; P = 0.251; table 3).

Patients in the muscular tissue oxygen saturation–guided care group had less moderate-to-severe nausea during the first postoperative 2 h (5% vs. 9%; risk ratio, 0.54; 95% CI, 0.32 to 0.91; P = 0.019) and 24 h (16% vs. 21%; risk ratio, 0.74; 95% CI, 0.55 to 0.99; P = 0.042); they reported a better quality of recovery in the following aspects: being able to enjoy food, feeling rested, feeling comfortable and in control, having a feeling of general well-being, and feeling well overall; they also reported a better quality of the first night’s sleep; however, these between-group differences were not significant after adjustment for multiple testing (table 3). The incidences of postoperative nausea and vomiting at very early (0 to 2 h), early (2 to 6 h), and late (6 to 24 h) postoperative stages were similar between groups (table 3). The severity of pain, the time to first ambulation, and the recovery of oral feeding tolerance were similar between groups (table 3). No patient in either group died or had any major organ system complications within 30 days after surgery.

Based on the per-protocol population, the incidence rates of 24-h postoperative nausea and vomiting were 30% (112 of 370) in the muscular tissue oxygen saturation–guided care group and 36% (139 of 383) in the usual care group, which was not statistically significantly different (risk ratio, 0.83; 95% CI, 0.68 to 1.02; P = 0.080; table S1 to S3 in Supplemental Digital Content, https://links.lww.com/ALN/C353). The effects of muscular tissue oxygen saturation–guided care on the incidence of 24-h postoperative nausea and vomiting in the modified intention-to-treat population (risk ratio, 0.94; 95% CI, 0.74 to 1.16; P = 0.553; table S4 in Supplemental Digital Content, https://links.lww.com/ALN/C353) and the per-protocol population (risk ratio, 0.85; 95% CI, 0.66 to 1.08; P = 0.189; table S5 in Supplemental Digital Content, https://links.lww.com/ALN/C353) remained insignificant after multivariable adjustments.

In patients with a body mass index of 25 or higher, the incidence of 24-h postoperative nausea and vomiting was significantly reduced from 41% (66 of 160) in the usual care group to 24% (44 of 183) in the muscular tissue oxygen saturation–guided care group (risk ratio, 0.58; 95% CI, 0.42 to 0.80; P < 0.001); this between-group difference remained significant after adjustment for multiple testing (fig. 3 and table S6-S8 in Supplemental Digital Content, https://links.lww.com/ALN/C353). The effect of muscular tissue oxygen saturation–guided care on the incidence of 24-h postoperative nausea and vomiting in patients with a body mass index of 25 or higher remained significant after multivariable adjustments (risk ratio, 0.58; 95% CI, 0.37 to 0.86; P = 0.006; table S9 in Supplemental Digital Content, https://links.lww.com/ALN/C353).

In patients with a body mass index of 25 or higher, muscular tissue oxygen saturation–guided care significantly reduced the incidence rates of 24-h postoperative nausea and vomiting from 33% (13 of 40) in the usual care group to 13% (7 of 56) in the muscular tissue oxygen saturation–guided care group in Peking University Third Hospital (risk ratio, 0.38; 95% CI, 0.17 to 0.88; P = 0.018) and from 68% (26 of 38) in the usual care group to 44% (16 of 36) in the muscular tissue oxygen saturation–guided care group in Shandong Provincial Hospital (risk ratio, 0.65; 95% CI, 0.42 to 0.99; P = 0.039), whereas in the other four hospitals, muscular tissue oxygen saturation–guided care did not lead to a significant between-group difference in the incidence of 24-h postoperative nausea and vomiting (fig. S2 in Supplemental Digital Content, https://links.lww.com/ALN/C353).

In this multicenter trial conducted in relatively young and healthy female patients having laparoscopic hysterectomy, we compared the effects of muscular tissue oxygen saturation–guided care versus usual care on postoperative nausea and vomiting. The goal of intervention was to maintain flank and forearm muscular tissue oxygen saturation greater than or equal to the baseline level or 70%, whichever was higher. Muscular tissue oxygen saturation–guided care did not significantly reduce the incidence of 24-h postoperative nausea and vomiting compared with usual care; however, it significantly reduced the incidence of 24-h postoperative nausea and vomiting in patients with a body mass index of 25 or higher based on preplanned subgroup analyses.

Compared with previous studies on goal-directed therapy, our study introduced the use of below-goal area under the curve to assess the degree of how well the predetermined goal is accomplished. This measure is important because if there is no between-group difference in area under the curve (i.e., the goal is not accomplished), it is difficult to understand the relationship between the goal-directed therapy and the outcome. Our findings suggested that the intervention protocol we used in this trial is effective in treating muscular tissue oxygen saturation decrements, because both the flank and forearm muscular tissue oxygen saturation areas under the curve in the muscular tissue oxygen saturation–guided care group were significantly smaller than those in the usual care group. The muscular tissue oxygen saturation–guided intervention protocol we used in this trial has prominent differences from the previously published ones.^27,28  One important difference is the use of a hemodynamic monitor in our protocol to facilitate the differential diagnosis and treatment of muscular tissue oxygen saturation decrements. This approach is related to the complicated physiology underlying tissue oxygen saturation measured by near-infrared spectroscopy. The factors that are frequently responsible for an adverse muscular tissue oxygen saturation decrease in anesthetized and paralyzed surgical patients are reduced cardiac output and reduced tissue perfusion.^14,29  Therefore, it is important to incorporate a hemodynamic monitor that offers dynamic information on intravascular volume, cardiac output, stroke volume, and systemic vascular resistance.

Our study presents a potential innovative measure for further quality improvement in perioperative care. Our patients were relatively young and healthy and underwent a relatively low-risk surgical procedure. We do not normally expect major perioperative complications in this patient population; rather, the priority is to promote enhanced recovery after surgery. Postoperative nausea and vomiting are roadblocks to enhanced recovery and have a relatively high incidence in this patient population.¹  In our study, despite the use of different prophylactic measures against postoperative nausea and vomiting, including propofol-based intravenous anesthesia, dexamethasone, and 5-HT3 antagonists, postoperative nausea and vomiting still occurred in 36% of patients in the usual care group. Although muscular tissue oxygen saturation–guided care did not significantly reduce the 24-h postoperative nausea and vomiting among all patients, it significantly reduced the postoperative nausea and vomiting risk in patients with a body mass index of 25 or higher. In patients with a body mass index of 25 or higher, muscular tissue oxygen saturation–guided care significantly reduced the occurrence of postoperative nausea and vomiting in two hospitals, but not the other four hospitals, suggesting the favorable outcome is primarily driven by the data from these two hospitals (fig. S2 in Supplemental Digital Content, https://links.lww.com/ALN/C353). The validity and mechanisms of this finding merit further investigation.

Muscular tissue oxygen saturation–guided care targets essential physiology (i.e., the balance between tissue oxygen consumption and supply) and offers personalized care when patient baseline measurements are referenced in decision-making. However, whether muscular tissue oxygen saturation–guided care reduces complications related to suboptimal tissue perfusion and oxygenation and promotes enhanced recovery after surgery remains to be further studied. Its values in different surgical patient populations having different surgical procedures remain to be elucidated. Muscular tissue oxygen saturation–guided care incurs certain costs and demands extra effort at the beginning of its implementation. Whether this personalized goal-directed muscular tissue oxygen saturation–guided intraoperative care is cost-effective remains unknown.

Our trial has certain limitations. Obviously, anesthesiologists could not execute muscular tissue oxygen saturation–guided care in a blinded manner. The lack of blinding may have introduced bias in the documentation of intraoperative information; however, the outcome data would be insulated from this source of bias because of the separation of these two processes. We did not directly measure gastrointestinal tissue oxygenation in this trial; therefore, whether the optimization of muscular tissue oxygenation also improves gastrointestinal tissue oxygenation remains a matter of speculation.⁷  We performed baseline measurements in awake patients in the operating room. Situation anxiety, preoperative fasting, and bowel preparation may enhance or reduce cardiac performance and tissue perfusion and oxygenation. Although we made efforts to obtain “real” baseline values, including administering anxiolytics to seemingly anxious patients and checking the vital signs measured outside of the operating room, the measurements obtained in the operating room may not always represent a patient’s resting and optimized condition. Finally, we may have underpowered this study by overestimating the effect size (20%) when planning the trial.

In conclusion, in relatively young and healthy female patients having laparoscopic hysterectomy, muscular tissue oxygen saturation–guided care during surgery is effective in treating muscular tissue oxygen saturation decrements but does not significantly reduce the incidence of 24-h postoperative nausea and vomiting compared with usual care. This innovative, personalized, goal-directed and muscular tissue oxygen saturation–guided pathway of care warrants further studies.

The authors thank CAS Medical Systems, Inc., (Branford, Connecticut; acquired by Edwards Lifesciences, Irvine, California) for providing the FORE-SIGHT ELITE tissue oximeters and disposable probes, and LiDCO (London, United Kingdom), for providing the hemodynamic monitors at no cost.

Supported by CAS Medical Systems, Inc., which funded two iMODIPONV research meetings that were conducted in Beijing, China, in August 2018 and January 2019. Also supported by institutional and/or departmental sources at Yale University (New Haven, Connecticut), Peking University (Beijing, China), Shandong Provincial Hospital (Jinan, China), Zhengzhou University (Zhengzhou, China), Hebei Medical University (Shijiazhuang, China), and Capital Medical University (Beijing, China).

Dr. Meng was a consultant to CAS Medical Systems, Inc. (Branford, Connecticut; now acquired by Edwards Lifesciences, Irvine, California). Dr. Dai was supported in part by CTSA grant No. UL1 TR001863 from the National Center for Advancing Translational Sciences (Bethesda, Maryland); however, the contents of this work are solely the responsibility of the authors and do not necessarily represent the official view of National Institutes of Health. The other authors declare no competing interests.

Full protocol available at: lingzhong.meng@yale.edu. Raw data available at: lingzhong.meng@yale.edu.