Abstract
Background
Many U.S. institutions have adopted postsurgical opioid-prescribing guidelines to
standardize prescribing practices, and yet there is inherent variability in patients’
opioid consumption after surgery. The utility of these guidelines is limited by the
fact that some patients’ needs will inevitably exceed them, and yet there are no evidence-based
tools to help providers identify these patients. In this study we aimed to maximize
the value of these guidelines by training machine learning models to predict patients
whose needs will be met by these smaller recommended prescriptions, and patients who
may require an additional degree of personalization. The aim of the present study
was to develop predictive models for determining whether a surgical patient's postdischarge
opioid requirement will fall above or below common opioid prescribing guidelines.
Methods
We conducted a retrospective cohort study of surgical patients at one institution
from 2017 to 2018. Patients were called after discharge to collect opioid consumption
data. Machine learning models were used to identify outlier opioid consumers (ie,
exceeding our institutional prescribing guidelines) using diagnosis codes, medical
history, in-hospital opioid use, and perioperative factors as predictors. External
validation was performed on opioid consumption data collected at a second institution
from 2020 to 2021, and sensitivity analysis was performed using a third institution’s
prescribing guidelines.
Results
The development and external validation cohorts included 1,867 and 498 patients, respectively.
Age, body mass index, tobacco use, preoperative opioid exposure, and in-hospital opioid
consumption were the strongest predictors of postdischarge consumption. A lasso regression
model exhibited an area under the receiver operating characteristic curve of 0.74
(95% confidence interval 0.67–0.81) in predicting postdischarge opioid consumption.
External validation of a limited lasso model yielded an area under the receiver operating
characteristic curve of 0.67 (0.60–0.74). Performance was preserved when evaluated
on another institution’s guidelines (area under the receiver operating characteristic
curve 0.76 [0.72–0.80]).
Conclusion
Patient characteristics reliably predict postdischarge opioid consumption in relation
to prescribing guidelines for both opioid-naive and exposed populations. This model
may be used to help providers confidently follow prescribing guidelines for patients
with typical opioid responsiveness and correctly pursue more personalized prescribing
for others.
To read this article in full you will need to make a payment
Purchase one-time access:
Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online accessOne-time access price info
- For academic or personal research use, select 'Academic and Personal'
- For corporate R&D use, select 'Corporate R&D Professionals'
Subscribe:
Subscribe to SurgeryAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
References
- COVID-19 and the drug overdose crisis: uncovering the deadliest months in the United States, January‒July 2020.Am J Public Health. 2021; 111: 1284-1291
- Vital statistics rapid release: provisional drug overdose data.https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htmDate accessed: September 29, 2021
- Causes and consequences of inadequate management of acute pain.Pain Med. 2010; 11: 1859-1871
- Description and impact of a comprehensive multispecialty multidisciplinary intervention to decrease opioid prescribing in surgery.Ann Surg. 2019; 270: 452-462
- Reduction in opioid prescribing through evidence-based prescribing guidelines.JAMA Surg. 2018; 153: 285-287
- Guideline for discharge opioid prescriptions after inpatient general surgical procedures.J Am Coll Surg. 2018; 226
- Guidelines for patient-centeredopioid prescribing and optimal FDA-compliant disposal of excess pills after inpatient operation: prospective clinical trial.J Am Coll Surg. 2021; : 232
- Opioid-prescribing guidelines for common surgical procedures: an expert panel consensus.J Am Coll Surg. 2018; 227: 411-418
- Probability of opioid prescription refilling after surgery: does initial prescription dose matter?.Ann Surg. 2018; 268: 271-276
- Defining optimal length of opioid pain medication prescription after common surgical procedures.JAMA Surg. 2018; 153: 37-43
- Comparing clinician consensus recommendations to patient-reported opioid use across multiple hospital systems.Ann Surg. 2020; 275: e361-e365
- Opioids prescribed after low-risk surgical procedures in the United States, 2004–2012.JAMA. 2016; 315: 1654-1657
- Opioid prescribing after surgery in the United States, Canada, and Sweden.JAMA Netw Open. 2019; 2e1910734
- Preoperative opioid use is independently associated with increased costs and worse outcomes after major abdominal surgery.Ann Surg. 2017; 265: 695-701
- Effect of pre-operative opioid exposure on surgical outcomes in elective laparoscopic cholecystectomy.Am J Surg. 2022; 223 (Published online June 24): 764-769
- Opioid prescribing recommendations.https://michigan-open.org/prescribing-recommendations/Date accessed: September 6, 2021
- Clinical Query 2.https://cq2.bidmc.org/home/Date accessed: September 6, 2021
- Data resources.https://www.cdc.gov/drugoverdose/resources/data.htmlDate accessed: September 6, 2021
- Opioid consumption after gender-affirming mastectomy and two other breast surgeries.J Surg Res. 2020; : 251
Kennedy CJ, Marwaha JS, Scalise PN, et al. Nonresponse adjustment using clinical and perioperative patient characteristics is critical for understanding post-discharge opioid consumption. bioRxiv. Published online July 7, 2021. https://doi.org/10.1101/2021.07.02.21259901. Accessed April 28, 2022.
- Reducing postoperative opioid pill prescribing via a quality improvement approach.Int J Qual Health Care. 2021; 33
- Defining adult overweight & obesity.https://www.cdc.gov/obesity/adult/defining.htmlDate accessed: September 6, 2021
- How to obtain the P value from a confidence interval.BMJ. 2011; 343: d2304
- scikit-learn.http://scikit-learn.sourceforge.netDate accessed: September 6, 2021
Chen T, Guestrin C. XGBoost. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM; 2016. https://doi.org/10.1145/2939672.2939785
- Association between initial opioid prescribing patterns and subsequent long-term use among opioid-naïve patients: a statewide retrospective cohort study.J Gen Intern Med. 2017; 32: 21-27
- Postsurgical prescriptions for opioid naive patients and association with overdose and misuse: retrospective cohort study.BMJ. 2018; 360: j5790
- Key substance use and mental health indicators in the United States.https://www.samhsa.gov/data/sites/default/files/reports/rpt29393/2019NSDUHFFRPDFWHTML/2019NSDUHFFR1PDFW090120.pdfDate accessed: December 9, 2021
- Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating.Springer International Publishing, 2019
- Transparent reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): the TRIPOD statement.Ann Intern Med. 2015; 162: 55-63
- Predicting postoperative opioid use with machine learning and insurance claims in opioid-naïve patients.Am J Surg. 2021; 222: 659-665
- Systematic evaluation of state policy interventions targeting the US opioid epidemic, 2007–2018.JAMA Netw Open. 2021; 4e2036687
- Association of state opioid duration limits with postoperative opioid prescribing.JAMA Netw Open. 2019; 2e1918361
- Patient-centered opioid prescribing: breaking away from one-size-fits-all prescribing guidelines.J Surg Res. 2021; 264: 1-7
- Association of opioid prescribing with opioid consumption after surgery in Michigan.JAMA Surg. 2019; 154e184234
- Results of a prospective, multicenter initiative aimed at developing opioid-prescribing guidelines after surgery.Ann Surg. 2018; 268: 457-468
Supplementary References
- Assessing the performance of prediction models: a framework for traditional and novel measures.Epidemiology. 2010; 21: 128-138
- Missing data should be handled differently for prediction than for description or causal explanation.J Clin Epidemiol. 2020; 125: 183-187
Article info
Publication history
Published online: May 05, 2022
Accepted:
March 20,
2022
Footnotes
Justin K. Yu and Jayson S. Marwaha are co-first authors.
Identification
Copyright
© 2022 Elsevier Inc. All rights reserved.
