Effect of the preoperative physical status on postoperative nausea and vomiting risk: a matched cohort study

In this study, we utilized a retrospective, matched cohort analysis to determine the relationship between preoperative medical status, assessed by ASA-PS, and the risk of PONV and PV development. Patients with ASA-PS 3 and higher showed a lower risk of PONV development than those with ASA-PS 1 and 2 (24% vs. 14%, respectively). However, there was no significant difference in the risk of PV. Our findings suggest that the patient’s physiological condition prior to surgery affected postoperative outcomes, particularly PONV occurrence. The matching algorithm resulted in a good balance across the studied groups in terms of patients, anesthesia, procedures, and medications (Tables 1, 2, and 3 [original data] and Supplementary files 1, 2 and 3 [sensitivity analysis]). Using risk factors with positive clinical evidence as covariates, we performed sensitivity analysis after considering that non-disease patients with obesity, smoking, and drinking may be included in ASA-PS 2 or 3. In sensitivity analysis, patients with ASA-PS 3 and higher showed a 32% and 14% lower risk of PONV development than did those with ASA-PS 1 and 2, respectively, and patients with ASA-PS 3 and higher did not have a significant difference in PV risk with did those with ASA-PS 1 and 2. The overall results revealed that the PONV risk in patients with ASA-PS 3 and higher was lower than that in patients with ASA-PS 1 and 2, while the risk of PV was not significantly different. The high ASA-PS classification used more intraoperative monitoring and vasopressor or inotropes than the low ASA-PS classification before and after matching. The results showed that, controlling similar perioperative factors at high ASA-PS scores, there may be many patients with conditions that require more intraoperative monitoring and the use of vasopressors or inotropes.

Previous studies included a better general condition as a risk factor of PONV (Kim et al. 2020a; Kwon et al. 2020; Cohen et al. 1994; Koivuranta et al. 1997). However, in one study, researchers examined PONV 72 h after surgery but failed to consider event occurrence time and postoperative factors despite the long observation time (Cohen et al. 1994). In another study, the PONV incidence pattern according to time was similar to our study, but no statistical analysis was performed (Koivuranta et al. 1997). Furthermore, previous researchers have not investigated PONV with the ASA-PS score as the primary risk factor. Comparisons were made between ASA-PS 1 and other scores or between ASA-PS 1–2 and ASA-PS 3 or higher; however, no comparisons were made between ASA-PS 2 or 3 and higher ASA-PS scores.

We found that patients with ASA-PS 3 and higher had a lower risk of developing PONV than ASA-PS 1 to 2, which was consistent with the findings from previous studies (Cohen et al. 1994; Koivuranta et al. 1997). However, according to whether non-disease patients with obesity, smoking, and drinking are included in ASA-PS 2 or 3, there were statistical differences in the association between ASA-PS classification. In this study, the ASA-PS classification used in sensitivity analysis may be more suitable for a recent version. In sensitivity analysis, a high ASA-PS classification score had a lower risk of PONV than a low ASA-PS classification score. These results showed consistency of the association between ASA-PS classification and PONV.

Vomiting is more fatal than nausea. When PV was analyzed separately, most results showed that PV was not associated with ASA-PS class. There were different results only between ASA-PS 1 and 2. However, when sensitivity analysis was performed, the results showed consistency of no association between ASA-PS class and PV. In particular, because smoking and obesity are factors known to reduce PONV risk (Sinclair et al. 1999; Sweeney 2002), they may have affected the differences. The differences in the association between ASA-PS classification and PONV and between ASA-PS classification and PV may be due to the effect of ASA-PV on postoperative nausea. Nausea is more common than vomiting. Considering the difference between nausea and vomiting, there may be a difference in the association with ASA-PS (Apfel et al. 2002b; Knapp and Beecher 1956). Nausea is a subjective sensation that must be assessed by the patient, not the observer. Nausea is best described as a desire to vomit without expelling muscle movement. As a subjective sensation, nausea should be considered a conscious cortical activity that may not affect the brainstem. Vomiting is a brainstem reflex and is not necessarily an exacerbated form of nausea, as nausea can occur without vomiting. Therefore, vomiting should be assessed independently. However, since nausea is difficult to distinguish and is frequently correlated with vomiting, vomiting without nausea is rare and may be a potential emetic phenomenon. In our study, nausea was evaluated as a result of inclusion in PONV (Gan 2006; Apfel et al. 2002b; Knapp and Beecher 1956).

Some preoperative conditions of patients that affect ASA-PS influence the choice of anesthetic agents, the pharmacokinetics of the medications used, the postoperative management, and the development of complications. In patients with increased intracranial pressure, intravenous anesthetic agents, such as propofol, are preferred to inhalation anesthetic agents (Butterworth et al. 2018). Etomidate is used in patients with hemodynamic instability. In short surgeries, etomidate increases PV (St Pierre et al. 2000). Sevoflurane may be associated with nephrotoxicity related to its fluoride content, although there is no definitive evidence. Thus, some anesthesiologists may not prefer sevoflurane in patients with kidney disease (Butterworth et al. 2018). In obese patients, highly lipophilic drugs, such as fentanyl and sufentanil, accumulate in fatty tissues when they are administered via infusion over long periods. Usually, highly lipophilic drugs show a significant increased volume of distribution in obese patients, and it seems that the dosing of these medications should be based on total body weight. However, because the majority of these drugs can accumulate in adipose tissues over time, prolonged effects can be seen (Hines and Jones 2021). Reduction in stroke volume can lead to protracted redistribution of opioids to the liver. This results in prolonged metabolization and lesser inactivation over time, followed by an increase in the duration of effects (Freye and Levy 2004).

We used risk factors with positive clinical evidence as covariates and found that most of them were related to anesthesia and surgery. Inhalation anesthetics, including N2O, are well-known risk factors of PONV. Anesthesiologists may attempt to use the least amount of anesthetics as possible for maintaining blood pressure in patients by using vasopressor or inotrope. Because higher ASA-PS classification used more vasopressor or inotrope, less anesthetics may be used in patients with higher ASA-PS classification. Patients with reduced levels of consciousness due to trauma or cerebrovascular injury, or patients with neurodegenerative changes due to dementia or other cognitive conditions, have reduced anesthesia requirements (Aranake et al. 2013). Hypoxia, acute metabolic acidosis, and acute hemorrhagic hypotension all cause a reduction of approximately 10 to 50% in the initial monitored anesthesia care (MAC) (Eger et al. 1965). However, drugs used for anesthesia and postsurgical recovery have a short duration of action or a short context-sensitive half time (Butterworth et al. 2018; Bailey 1997; Egan 1995). In this study, over 75% of PONV occurred 2 h post-anesthetic and over 50% occurred 5 h post-anesthetic. Therefore, the effects may be limited and not evident from surgery to the recovery room. In addition, our findings suggest that other risk factors not included in this study should be considered for PONV developing. More than three-quarters of PONV occurred 2 h post-anesthetic. In cases of PONV occurring 2 h post-anesthetic, antibiotic, and nonsteroidal anti-inflammatory drug use, the severity of postoperative pain, the postoperative fasting period, and the oxygen supply must be considered. Patients with high ASA-PS scores who have asthma or renal disease may have difficulty with nonsteroidal anti-inflammatory drugs, those with cardiopulmonary disease may require prolonged oxygen therapy, and those with aspiration risk may require prolonged fasting periods. The postsurgical role of these factors remains controversial (Gan 2006; Gan et al. 2019; Gan et al. 2014; Junger et al. 2001; Kearney et al. 1998; Stadler et al. 2003; Tramer 2001), but the association between ASA-PS class and these factors also needs to be considered.

The strength of this study was that a large number of patients were analyzed. Considering that there may be a difference between vomiting and nausea, the analysis of PV was separated. However, this study has several limitations. First, the reliability of the ASA-PS Classification System should be considered. ASA-PS is the most commonly used tool to classify the preoperative condition. It was created in 1941, revised in 1961, and revised again in 2014 (Mayhew et al. 2019). Although definitions and examples are described for ASA-PS, there are bound to be limitations in classifying all patients by these definitions and examples, and differences may arise due to these limitations. Therefore, it is difficult to maintain consistency in classification; this has already been reported in several studies (Haynes and Lawler 1995; Owens et al. 1978). We included patients from 2015 by considering the 2014 revision. In our study, 15% of the total subjects should have belonged to ASA-PS class 2 or 3 because of smoking, obesity (ASA-PS 2, 30 ≤ BMI < 40; ASA-PS 3, 40 ≤ BMI), and social alcohol consumption; however, they belonged to ASA-PS 1. We performed a sensitivity analysis for reducing this uncertainty. In addition, we found that there may be a difference in the number of vulnerable patients according to ASA-PS classification through secondary outcomes (including intraoperative invasive monitoring and use of vasopressor or inotrope). A second limitation was in the method of data collection. Patients complain of PONV much more often than indicated in the chart (Cohen et al. 1994). Direct, specific questions report actual PONV much more than spontaneous patient reports (Apfel et al. 1998; Apfel et al. 2002c), and higher workloads of nurses or physicians may result in fewer vomiting events being reported (Sinclair et al. 1999). The nature and severity of PONVs collected in studies may affect the accuracy or applicability of independent risk factor detection (Gan 2006). Third, despite the importance of the effects of additional antiemetics on PONV, we used antiemetics as a binary variable. It is well known that additional antiemetics are effective against PONV. However, because our data concerning the history of PONV and details about antiemetics, such as the timing and number of administration, was insufficient, we could not analyze the effect of the additional use of antiemetics according to PONV risk. Additional antiemetics could affect the results, and it seems necessary to analyze the effects of the controlled use of additional antiemetics in a further study.

In conclusion, in an analysis that controlled for PONV-related anesthesia and surgical factors, ASA-PS 3 and higher was found to reduce the risk of PONV but was not related to PV. Despite the limitations, the findings of this study may help anesthesiologists determine the risk of PONV in patients undergoing surgery under anesthesia. Considering the times of PONV occurrence that we observed which suggest that the effects of anesthesia and surgery factors may be limited, the effects of postoperative factors on PONV may need to be considered. Future studies focusing on the association of PONV with preoperative physical status will be needed.

留言 (0)

沒有登入
gif