Cannabis use is associated with a small increase in the risk of postoperative nausea and vomiting: a retrospective machine-learning causal analysis

Cannabis smoking and abdominal wall reconstruction outcomes: a propensity score-matched analysis

PURPOSE

Despite increasing use of cannabis, literature on perioperative effects is lagging. We compared active cannabis-smokers versus non-smokers and postoperative wound morbidity and reoperations following open abdominal wall reconstruction (AWR).

METHODS

Patients who underwent open, clean, AWR with transversus abdominis release and retromuscular synthetic mesh placement at our institution between January 2014 and May 2022 were identified using the Abdominal Core Health Quality Collaborative database. Active cannabis-smokers were 1:3 propensity matched to non-smokers based on demographics and comorbidities. Wound complications, 30 day morbidity, pain (PROMIS 3a-Pain Intensity), and hernia-specific quality of life (HerQles) were compared.

RESULTS

Seventy-two cannabis-smokers were matched to 216 non-smokers. SSO (18% vs 17% p = 0.86), SSI (11.1% vs 9.3%, p = 0.65), SSOPI (12% vs 12%, p = 0.92), and all postoperative complications (46% vs 43%, p = 0.63) were similar between cannabis-smokers and non-smokers. Reoperations were more common in the cannabis-smoker group (8.3% vs 2.8%, p = 0.041), driven by major wound complications (6.9% vs 3.2%, p = 0.004). No mesh excisions occurred. HerQles scores were similar at baseline (22 [11, 41] vs 35 [14, 55], p = 0.06), and were worse for cannabis-smokers compared to non-smokers at 30 days (30 [12, 50] vs 38 [20, 67], p = 0.032), but not significantly different at 1 year postoperatively (72 [53, 90] vs 78 [57, 92], p = 0.39). Pain scores were worse for cannabis-smokers compared to non-smokers at 30 days postoperatively (52 [46, 58] vs 49 [44, 54], p = 0.01), but there were no differences at 6 months or 1 year postoperatively (p > 0.05 for all).

CONCLUSION

Cannabis smokers will likely experience similar complication rates after clean, open AWR, but should be counseled that despite similar wound complication rates, the severity of their wound complications may be greater than non-smokers.

Prediction of Complications and Prognostication in Perioperative Medicine: A Systematic Review and PROBAST Assessment of Machine Learning Tools

BACKGROUND

The utilization of artificial intelligence and machine learning as diagnostic and predictive tools in perioperative medicine holds great promise. Indeed, many studies have been performed in recent years to explore the potential. The purpose of this systematic review is to assess the current state of machine learning in perioperative medicine, its utility in prediction of complications and prognostication, and limitations related to bias and validation.

METHODS

A multidisciplinary team of clinicians and engineers conducted a systematic review using the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) protocol. Multiple databases were searched, including Scopus, Cumulative Index to Nursing and Allied Health Literature (CINAHL), the Cochrane Library, PubMed, Medline, Embase, and Web of Science. The systematic review focused on study design, type of machine learning model used, validation techniques applied, and reported model performance on prediction of complications and prognostication. This review further classified outcomes and machine learning applications using an ad hoc classification system. The Prediction model Risk Of Bias Assessment Tool (PROBAST) was used to assess risk of bias and applicability of the studies.

RESULTS

A total of 103 studies were identified. The models reported in the literature were primarily based on single-center validations (75%), with only 13% being externally validated across multiple centers. Most of the mortality models demonstrated a limited ability to discriminate and classify effectively. The PROBAST assessment indicated a high risk of systematic errors in predicted outcomes and artificial intelligence or machine learning applications.

CONCLUSIONS

The findings indicate that the development of this field is still in its early stages. This systematic review indicates that application of machine learning in perioperative medicine is still at an early stage. While many studies suggest potential utility, several key challenges must be first overcome before their introduction into clinical practice.

EDITOR’S PERSPECTIVE

Cannabis Users Require More Anesthetic Agents for General Anesthesia in Ambulatory Oral and Maxillofacial Surgery Procedures

BACKGROUND

The use of cannabis in the general population has steadily increased over recent years and there is limited literature regarding the anesthetic implications of chronic cannabis use, particularly in the setting of outpatient anesthesia.

PURPOSE

To determine whether chronic cannabis users undergoing deep sedation or general anesthesia during ambulatory procedures require more anesthetic agents than nonusers.

STUDY DESIGN, SETTING, AND SAMPLE

A retrospective cohort study of subjects undergoing deep sedation or general anesthesia at the Oklahoma University Oral and Maxillofacial Surgery Clinic from January to December 2022 was performed. The inclusion criteria were duration of anesthetic procedure between 15 to 40 minutes, use of propofol, fentanyl, ketamine, and midazolam, and extraction of at least 2 teeth. The exclusion criterion was patients undergoing adjunctive procedures other than extractions during sedation.

PREDICTOR VARIABLE

Cannabis use status was grouped as users and nonusers. A user was defined as a subject who self-reported any regular use of cannabis.

OUTCOME VARIABLE

The primary outcome variable was the amount of intravenous anesthetic agents administered. Secondary outcome variables included the length of the procedure and the number of teeth extracted.

COVARIATES

Age, sex, and the senior-most resident involved in the sedation.

ANALYSES

IBM SPSS was utilized to perform descriptive statistics, paired t-tests, ANOVA, and multivariate linear regression. A level of significance of 5% (P < .05) was used for all analyses.

RESULTS

Four hundred and ninety nine subjects were identified, 189 met the inclusion criteria, and 57 reported using cannabis. The mean age of nonusers was 28.2 ± 7.8 years and that of users was 26.6 ± 6.4 years (P = .09). Females represented 71.9% of nonusers and 72.7% of users. Cannabis users received significantly more propofol (117.5 mg ± 71.3 vs 152.5 mg ± 101.8; P = .004), midazolam (4.7 mg ± 1.0 vs 5.1 mg ± 1.5; P = .01), ketamine (40.2 mg ± 15.7 vs 46.1 mg ± 16.9; P = .01), and fentanyl (75.2 μg ± 26.3 vs 88.6 μg ± 32.8; P = .002) than nonusers, despite extracting a similar number of teeth (4.5 ± 3.1 vs 4.4 ± 3.5; P = .37) in a similar amount of time (25.5 ± 7.3 vs 27.3 ± 7.8; P = .06).

CONCLUSION AND RELEVANCE

Cannabis users required more propofol, midazolam, ketamine, and fentanyl than non-cannabis users during outpatient oral and maxillofacial surgery procedures.

ASRA Pain Medicine consensus guidelines on the management of the perioperative patient on cannabis and cannabinoids

BACKGROUND

The past two decades have seen an increase in cannabis use due to both regulatory changes and an interest in potential therapeutic effects of the substance, yet many aspects of the substance and their health implications remain controversial or unclear.

METHODS

In November 2020, the American Society of Regional Anesthesia and Pain Medicine charged the Cannabis Working Group to develop guidelines for the perioperative use of cannabis. The Perioperative Use of Cannabis and Cannabinoids Guidelines Committee was charged with drafting responses to the nine key questions using a modified Delphi method with the overall goal of producing a document focused on the safe management of surgical patients using cannabinoids. A consensus recommendation required ≥75% agreement.

RESULTS

Nine questions were selected, with 100% consensus achieved on third-round voting. Topics addressed included perioperative screening, postponement of elective surgery, concomitant use of opioid and cannabis perioperatively, implications for parturients, adjustment in anesthetic and analgesics intraoperatively, postoperative monitoring, cannabis use disorder, and postoperative concerns. Surgical patients using cannabinoids are at potential increased risk for negative perioperative outcomes.

CONCLUSIONS

Specific clinical recommendations for perioperative management of cannabis and cannabinoids were successfully created.

Marijuana use does not affect weight loss or complication rate after bariatric surgery

INTRODUCTION

Marijuana use has been legalized in several states. It is unclear if marijuana use affects weight loss outcomes or complication rates following bariatric surgery. The purpose of this study was to determine if patients who use marijuana had higher complication rates or lower weight loss compared with non-users.

METHODS

All patients at a single institution who underwent primary bariatric surgery between July 2015 and July 2020 at a single institution after the legalization of marijuana within the jurisdiction were included. Data regarding marijuana use, weight and complications were abstracted retrospectively. Differences between groups were evaluated with Wilcoxon Rank-Sum tests and Fisher Freeman Halton test. Trends for marijuana use over time was evaluated with simple linear regression on summary data.

RESULTS

1107 patients met inclusion criteria. 798 (73.3%) were never users, 225 (19.4%) were previous users, and 84 (7.2%) were active users. The proportion of active users and previous users increased over time, with significantly more prior marijuana use reported in more recent years (p = 0.014). Active users had significantly higher pre-procedural BMIs than never users: 48.7 vs. 46.3 (p = 0.03). Any marijuana use (active and previous users) was associated with higher preoperative weight compared to never: 136.4 kg vs. 130.6 kg (p = 0.001). Overall complication rate was low in all groups, and there was no difference in the rates of any complications. Active and previous users tended to lose less weight than never users, but this was not statistically significant (p = 0.17).

CONCLUSIONS

Active and prior marijuana users tend to have higher BMIs on presentation, but use was not associated with complications or percent weight loss. The incidence of patient reported marijuana use is increasing in the study population. More studies on the effects of marijuana use in this patient population are warranted.

Machine learning in perioperative medicine: a systematic review

Background

Risk stratification plays a central role in anesthetic evaluation. The use of Big Data and machine learning (ML) offers considerable advantages for collection and evaluation of large amounts of complex health-care data. We conducted a systematic review to understand the role of ML in the development of predictive post-surgical outcome models and risk stratification.

Methods

Following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines, we selected the period of the research for studies from 1 January 2015 up to 30 March 2021. A systematic search in Scopus, CINAHL, the Cochrane Library, PubMed, and MeSH databases was performed; the strings of research included different combinations of keywords: “risk prediction,” “surgery,” “machine learning,” “intensive care unit (ICU),” and “anesthesia” “perioperative.” We identified 36 eligible studies. This study evaluates the quality of reporting of prediction models using the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) checklist.

Results

The most considered outcomes were mortality risk, systemic complications (pulmonary, cardiovascular, acute kidney injury (AKI), etc.), ICU admission, anesthesiologic risk and prolonged length of hospital stay. Not all the study completely followed the TRIPOD checklist, but the quality was overall acceptable with 75% of studies (Rev #2, comm #minor issue) showing an adherence rate to TRIPOD more than 60%. The most frequently used algorithms were gradient boosting (n = 13), random forest (n = 10), logistic regression (LR; n = 7), artificial neural networks (ANNs; n = 6), and support vector machines (SVM; n = 6). Models with best performance were random forest and gradient boosting, with AUC > 0.90.

Conclusions

The application of ML in medicine appears to have a great potential. From our analysis, depending on the input features considered and on the specific prediction task, ML algorithms seem effective in outcomes prediction more accurately than validated prognostic scores and traditional statistics. Thus, our review encourages the healthcare domain and artificial intelligence (AI) developers to adopt an interdisciplinary and systemic approach to evaluate the overall impact of AI on perioperative risk assessment and on further health care settings as well.

Effect of preoperative cannabis use on perioperative outcomes: a retrospective cohort study

INTRODUCTION

The reported use of cannabis within surgical population is increasing. Cannabis use is potentially associated with increased harms and varied effects on pain control. These have important implications to perioperative care.

METHODS

We conducted a retrospective cohort study comparing surgical patients reporting cannabis use preoperatively to control patients with no cannabis exposure, in a 1:2 ratio. To control for confounding, we used a propensity score-matched analysis to assess the adjusted association between cannabis use and study outcomes. Our primary outcome was a composite of (1) respiratory arrest or cardiac arrest, (2) intensive care admission, (3) stroke, (4) myocardial infarction and (5) mortality during this hospital stay. Secondarily, we assessed the effects on pain control, opioid usage, induction agent dose and nausea-vomiting.

RESULTS

Between January 2018 and March 2019, we captured 1818 patients consisting of cannabis users (606) and controls (1212). For propensity score-matched analyses, 524 cannabis patients were compared with 1152 control patients. No difference in the incidence of composite outcome was observed (OR 1.06, 95% CI 0.23 to 3.98). Although a higher incidence of arrhythmias (2.7% vs 1.6%) and decreased incidence of nausea-vomiting needing treatment (9.6% vs 12.6%) was observed with cannabis users vs controls, results were not statistically significant. No significant differences were observed with other secondary outcomes.

CONCLUSION

Our results do not demonstrate a convincing association between self-reported cannabis use and major surgical outcomes or pain management. Perioperative decisions should be made based on considerations of dose, duration, and indication.