Risk Prediction Model Can Help Determine, Curb Opioid Misuse

In response to the addiction and medical risks associated with excessive opioid use, Michigan Medicine researchers developed a risk prediction model that can help define appropriate opioid dosages for patients, according to a study published by the University of Michigan.

Over 90 percent of surgical patients in the US have been prescribed opioids after surgery, and of these patients, between 9 and 13 percent have no prior experience using them, the study stated.

Despite the benefits that opioids can provide following an operation, there are also various risks associated with their use.

“Opioid-based pain medications have an important therapeutic role in surgical recovery, but they also introduce risks of long-term use, physical dependence, and addiction. Many people use them for acute pain after surgery and stop without incident, but some do not,” said Anne Fernandez, PhD, an assistant professor of psychiatry and corresponding author on the study, in the press release.

Fernandez and his fellow researchers developed the risk prediction model to help identify and care for patients who might develop an addiction to opioids.

Researchers used a sample of derivation and validation cohorts from the Michigan Genomics Initiative. The model consisted of three different versions: a full model of 216 predictors, a restricted model of ten predictors, and a minimal model of five predictors.

All three performed better at predicting continued opioid use than existing models, the study shows.

Researchers also found that the three versions were more successful at predicting opioid use among preoperative opioid users than inexperienced users.

However, researchers noted that the minimal version of the model provided the worst evaluation among inexperienced opioid users.

Researchers concluded that the restricted model had the most practical sample of ten predictors and the ability to coincide with any setting. It also included patient-reported measures, which made it more accurate in determining opioid use among inexperienced opioid users.

“We chose the restricted, 10-variable model based on two things: parsimony and accuracy,” Fernandez said. “To be useful in a real-world setting, the model has to be simple enough to implement in clinical care. A model with 100s of variables is not very easy to implement…The 10-variable model was the simplest model that didn’t sacrifice accuracy.”

However, despite its success, one limitation in the restricted model is race. Researchers acknowledged that a lack of race data  might affect results.

In the future, researchers hope to implement this model in various settings to evaluate its performance further and see if it continues to be successful externally.

Outside of opioid use, risk prediction models have been proven reliable in various other use cases.

A study from March 2021 shows how risk prediction models can help decrease suicide rates.

Also, a risk prediction model can help determine the risk of breast cancer, a study from last February 2021. It does this by using data from mammograms and considering age and family history to determine if, when, and how often a patient should receive screening.

Further, risk prediction models have proved useful during the COVID-19 pandemic. A study from the pandemic's peak shows how predictive analytics can help clinicians evaluate the dangers of COVID-19, how it will affect healthcare resources and the type of care a particular patient may need.