New study reveals cannabis use during pregnancy more prevalent than previously thought – users need to be informed of risks
A groundbreaking study presented at ADLM 2024 has uncovered alarming rates of cannabis use during pregnancy, highlighting the need for improved screening and patient education.
Widespread cannabis use in pregnancy
Recent research conducted by a team from US-based NMS Labs has shed light on the prevalence of cannabinoid exposure in utero during the period from 2019 to 2023. The study, led by Dr Alexandria Reinhart, analysed 90,384 umbilical cord samples submitted for toxicology testing. The results were startling: 44% of the samples tested positive for at least one of approximately 60 analytes included in the testing panel. Even more concerning was the finding that cannabinoids accounted for 59%-63% of all positive results, making them the most commonly detected substance.
Dr Reinhart expressed her astonishment at the findings, saying: “The sheer amount of cannabinoid positivity we found in comparison to all the drugs that we run on our umbilical cord toxicity testing was pretty astounding.”
Implications for maternal and foetal health
The increasing legalisation of cannabis across the United States has led to a surge in recreational use. However, the effects of cannabis on the developing foetus are not yet fully understood. Current medical recommendations strongly advise against cannabis use and exposure during pregnancy due to its association with several adverse outcomes.
These negative consequences include:
- Preterm birth
- Foetal growth restriction
- Low birth weight
- Developmental deficits
The high prevalence of cannabis use during pregnancy uncovered by this study raises significant concerns about the potential long-term impacts on maternal and foetal health.
Call for increased vigilance and education
In light of these findings, Dr Reinhart emphasised the importance of clinical laboratories remaining vigilant in testing for cannabinoids in pregnant individuals. This increased screening would enable healthcare providers to identify at-risk patients and provide crucial education about the potential harm that cannabis can do to a developing foetus.
The research team suggests that improved screening methods and more comprehensive patient education programmes could help mitigate the risks associated with cannabis use during pregnancy. By identifying cannabis users early in their pregnancies, healthcare providers can intervene and offer support to help patients abstain from use for the duration of their pregnancy.
Machine learning model predicts post-surgical opioid use duration
In a related study also presented at ADLM 2024, researchers have developed a machine learning model that can predict the duration of opioid use following surgery. This innovative approach could revolutionise pain management strategies and help reduce the risk of opioid dependence and addiction.
Addressing variability in opioid response
Hydrocodone, the most commonly prescribed opioid in the United States, is frequently used for post-surgical pain management. However, there is significant variability in patients’ response to hydrocodone therapy, including the duration of time they require the drug to manage their pain effectively.
Dr Hunter Miller, along with colleagues from the University of Louisville and a researcher from ARUP Institute for Clinical and Experimental Pathology, set out to tackle this issue by developing machine learning models to predict postoperative hydrocodone use duration in patients who had undergone orthopaedic surgery.
Two predictive models
The research team developed two different models:
- A fast and frugal tree (FFTree) model
- An extreme gradient boosting approach (xgBoost) model
Both models incorporated a range of patient data, including demographics, genetic test results, concurrently prescribed medications, and other clinical laboratory test results.
To evaluate the models’ effectiveness, the researchers used them to predict the duration of hydrocodone use for 79 patients for whom they already had hydrocodone use duration information. The results were impressive, with both models demonstrating good to excellent performance when classifying patients as either “short” or “long” duration users.
The FFTree model achieved a sensitivity of 0.80 and a specificity of 0.76, while the xgBoost model demonstrated a sensitivity of 0.87 and a specificity of 0.63.
Potential for personalised pain management
Dr Miller highlighted the potential impact of these models on clinical practice, saying: “Currently, when it comes to pain management, most clinicians are kind of just taking a shot in the dark, because they don’t really know how a patient is going to respond to a drug.”
He went on to describe a possible future scenario where “a physician could theoretically put a patient’s information into the model, estimate a probability for how long a patient is going to be on hydrocodone, and potentially switch them to a different therapeutic strategy if they have a high risk of prolonged use.”
This approach could lead to more personalised pain management strategies, potentially reducing the risk of opioid dependence and improving patient outcomes.
Conclusion
The two studies presented at ADLM 2024 offer valuable insights into pressing issues in maternal health and pain management. The alarming prevalence of cannabis use during pregnancy underscores the need for improved screening and patient education, while the development of machine learning models to predict opioid use duration shows promise for more tailored and effective pain management strategies.
